DOLOMITIZATION AND POROSITY EVOLUTION OF MIDDLE BAKKEN

MEMBER, ELM COULEE FIELD AND FACIES CHARACTERIZATION,

CHEMOSTRATIGRAPHY AND ORGANIC-RICHNESS OF
UPPER BAKKEN SHALE, WILLISTON BASIN

by
Dipanwita Nandy
Copyright by Dipanwita Nandy 2017
All Rights Reserved
 A thesis submitted to the Faculty and the Board of Trustees of the Colorado School of
Mines in partial fulfillment of the requirements for the degree of Doctor of Philosophy (Geology).

Golden, Colorado

Date: _________________

Signed: ___________________

Dipanwita Nandy

Signed: ___________________

Dr. Stephen A. Sonnenberg

Thesis Advisor

Golden, Colorado

Date: _________________

Signed: ___________________

M. Stephen Enders
Professor and Head
Department of Geology and Geological Engineering

ii
 1. ABSTRACT

This thesis is focused on improving the current geological understanding of the Bakken Shale Play,
which currently produces with a rate of about 1 million BOPD and contributes about 12% of the total
indigenous oil production of the US. In the first part of the thesis, I investigate the effect of dolomitization
on the total porosity, pore architecture and pore-size of the Middle Bakken Member (MBM), which is one
of the major producing intervals of the Bakken Shale Play. The second part of this research is focused on
the organic-rich Upper Bakken Shale (UBS), which along with the Lower Bakken Shale, is the source rock
interval for the Bakken Petroleum System. However, the potential of UBS as a future production interval
is highlighted by the recent exploration successes and historical examples of production, which emphasizes
the need for a comprehensive characterization of the UBS with respect to its facies and internal stratigraphic
framework. I have investigated various aspects of the UBS, such as its facies types, their distribution in the
basin and sedimentary processes. Moreover, chemostratigraphy was used to provide an internal
stratigraphic framework for the UBS. The third part of this study is focused on understanding the role and
interplay of the factors that controlled the organic-richness in UBS.

The first part of this research, which is focused on the MBM of the Elm Coulee Field, is based on
the results of core description, thin section and scanning electron microscope (SEM) petrography, stable
isotope of carbon and oxygen, routine core analysis (RCA), X-Ray Diffraction (XRD) and Mercury
Intrusion Porosimetry (MIP). The relationships between dolomite content, total porosity, and pore size
suggested that the porosity evolution in the MBM is closely linked with the dolomitization process. An
increase in the dolomite content up to 60 wt.% results in an overall increase in the development of
intercrystalline pores, total porosity, pore size, and dolomite crystal size. An increase in dolomite content
above 60 wt.% marks overdolomitization of the MBM, which results in an increase in the size of dolomite
crystals, and a decrease in both total porosity and pore size. My findings suggest that MBM of the Elm
Coulee Field was dolomitized by seepage-reflux of the mesohaline brine. I propose that in the MBM, the
dolomitization process involved an open-system volume-for-volume replacement of precursor calcite by
dolomite, and the degree of dolomitization was dependent on the external supply of Mg2+and (CO3)2- ions.
Dolomitization by volume-to-volume replacement results in an increase in the pore size, while the total
porosity remains the same during the process. In MBM, an increase in porosity with an increase in dolomite
content resulted from a relatively less compaction of dolomite-rich intervals, in comparison to the calcite-
rich intervals, because the later is more susceptible to compaction. In the overdolomitization stage, dolomite
crystals form interlocking fabrics, which in turn resulted in a decrease in the total porosity and pore size
due to dolomite cement overgrowth.

iii
 I identify six lithofacies of the UBS from core description and thin-section petrographic study;
however, 95% of the UBS consists of the three main facies: siliceous mudstone, massive to finely laminated
silt-bearing mudstone, and macrofossil-bearing silt-rich mudstone. Three regionally correlatable
chemostratigraphic packages (bottom to top): Sub-unit 1a, 1b, and Unit-2 were identified in the UBS. My
study indicates that the maximum flooding surface is located in sub-unit 1b, near the contact of Unit-2 and
sub-unit 1b. The results of this research show that during the deposition of the UBS, the basin was semi-
restricted, had stratified water column, and the regeneration of the biolimiting nutrients such as P and N
was one of the primary sources of nutrients. Moreover, the redox condition, paleoproductivity, and the
influx of detrital and biogenic sediments varied during the deposition of the UBS, primarily due to a relative
change in the sea level. This resulted in a temporal variation in TOC and the deposition of four distinct sub-
units in the UBS (bottom to top): 1a, 1b, 2a, and 2b. Sub-units 1a and 1b were deposited in a strongly
euxinic condition, which existed in the bottom water and extended to the photic zone intermittently.
Whereas, during the deposition of sub-units 2a and 2b, the conditions were predominantly sub-oxic to
anoxic along the sediment-water interface. Moreover, paleoproductivity was high during the deposition of
sub-unit 1a and 1b, and paleoproductivity declined during the deposition of sub-units 2a and 2b.

iv
 TABLE OF CONTENTS

ABSTRACT ........................................................................................................................................... iii

LIST OF FIGURES ................................................................................................................................... viii
LIST OF TABLES ..................................................................................................................................... xxi
ACKNOWLEDGEMENTS ..................................................................................................................... xxiii
CHAPTER 1 INTRODUCTION .............................................................................................................. 1
1.1 Motivation .......................................................................................................................... 2
1.2 Research Objectives and Methods...................................................................................... 3
1.3 Organization of the Thesis.................................................................................................. 4
1.4 Conference Proceedings and Publication ........................................................................... 4
1.5 References .......................................................................................................................... 5
CHAPTER 2 DOLOMITIZATION AND ITS EFFECT ON POROSITY EVOLUTION IN MIDDLE
BAKKEN, ELM COULEE FIELD ................................................................................... 7
2.1 Introduction ........................................................................................................................ 8
2.2 Literature review on origin of porosity in dolostones ........................................................ 8
2.3 Geological Setting ............................................................................................................ 10
2.4 Methodology and Dataset ................................................................................................. 11
2.5 Results and Interpretation ................................................................................................. 14
2.5.1 Lithofacies Characterization .............................................................................. 15
2.5.2 Diagenesis .......................................................................................................... 16
2.5.3 Petrography and Stable Isotope of the Middle Bakken Dolomites .................... 25
2.5.4 Porosity Characterization ................................................................................... 26
2.5.5 Relationship between dolomite content and porosity ........................................ 30
2.6 Discussion ......................................................................................................................... 33
2.6.1 Dolomitization in Elm Coulee Field .................................................................. 33
2.6.2 Dolomitization process and porosity evolution ................................................. 37
2.6.3 Conceptual Model for Porosity Evolution ......................................................... 39
2.7 Conclusions ....................................................................................................................... 44
28 References ......................................................................................................................... 44
CHAPTER 3 FACIES CHARACTERIZATION AND CHEMOSTRATIGRAPHY OF ORGANIC-
RICH UPPER BAKKEN SHALE, WILLISTON BASIN ............................................... 50
3.1 Introduction ....................................................................................................................... 50
3.2 Geological Setting ............................................................................................................. 52
3.3 Methodology and Background .......................................................................................... 55
3.3.1 Core and Dataset ................................................................................................ 56

v
 3.3.2 ED-XRF Instrument Procedure and Interpretation Background ........................ 57
3.3.3 Detrital, Biogenic, and Redox Proxies ............................................................... 58
3.3.4 Core Description and Petrography ..................................................................... 60
3.3.5 SRA and ICPMS Measurements ........................................................................ 61
3.4 Result and Interpretation ................................................................................................... 61
3.4.1 Facies Description and Depositional Processes ................................................. 61
3.4.2 Factor Analysis and Element Relationships using XRF data............................. 77
3.4.3 Crossplots Between Elements concentrations.................................................... 77
3.4.4 XRD-XRF comparison ...................................................................................... 81
3.4.5 Elemental chemostratigraphic correlation ......................................................... 83
3.4.6 Regional distribution of Chemostratigraphic Units ........................................... 83
3.4.7 Distribution of Facies in Chemostratigraphic Units .......................................... 92
3.4.8 Gamma Ray Log and TOC Trends .................................................................... 92
3.5 Discussion ......................................................................................................................... 96
3.5.1 Detrital Sediment Influx .................................................................................... 97
3.5.2 Sedimentation Rate ............................................................................................ 97
3.5.3 Depth of Basin ................................................................................................... 98
3.5.4 Redox condition and bioturbation .................................................................... 100
3.5.5 Sequence Stratigraphic Interpretation .............................................................. 104
3.5.6 Mineralogy Based Resource Potential of UBS Chemostratigraphic Units ...... 107
3.6 Conclusions ..................................................................................................................... 108
3.7 References ....................................................................................................................... 108
CHAPTER 4 FACTORS CONTROLLING ORGANIC RICHNESS OF UPPER BAKKEN
SHALE, WILLISTON BASIN ...................................................................................... 117
4.1 Introduction ..................................................................................................................... 118
4.2 Methodology: Sampling and analytical procedures ........................................................ 120
4.2.1 Proxies for paleoredox condition ..................................................................... 123
4.2.2 C-S-Fe relationship .......................................................................................... 125
4.2.3 Proxy for Paleoproductivity and Organic Matter type ..................................... 127
4.3 Results and Interpretation ................................................................................................ 127
4.3.1 Chemostratigraphy, Core description, and Petrography .................................. 127
4.3.2 TOC, OM Characteristics and Type ................................................................ 128
4.3.3 Effect of Detrital and Biogenic Sediment Influx on OM Richness ................. 131
4.3.4 Redox Proxy: C-S-Fe relationship ................................................................... 135
4.3.5 TOC-Trace Elements Relationship .................................................................. 137

vi
 4.3.6 EF-Mo and EF-U relationship ......................................................................... 140
4.3.7 Mo-TOC Relationship and Basin Restriction .................................................. 141
4.3.8 Redox-condition and OM preservation ............................................................ 142
4.3.9 Proxy for Paleoproductivity ............................................................................. 143
4.4 Discussion ....................................................................................................................... 145
4.4.1 Depositional Setting ......................................................................................... 145
4.4.2 Model for Organic Richness in Chemostratigraphic Units of UBS ................. 148
4.5 Conclusion ....................................................................................................................... 152
4.6 References ....................................................................................................................... 153
CHAPTER 5 CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK .................... 159
5.1 Conclusion ....................................................................................................................... 159
5.2 Recommendations for Future Work ................................................................................ 160
APPENDIX A DIGITIZED CORE DESCRIPTION OF MIDDLE BAKKEN MEMBER, ELM
COULEE FIELD .......................................................................................................... 159
APPENDIX B RAW DATA AND SOFT COPIES ................................................................................ 170
APPENDIX C XRF MEASUREMENT QUALITY ASSURANCE AND QUALITY CHECK ........... 170
C.1 Sample Preparation......................................................................................................... 172
C.2 Optimal Detection Time ................................................................................................. 173
C.3 Reproducibility Test and XRF Measurement Validation ............................................... 175
C.4 Repeatability Test ........................................................................................................... 178
APPENDIX D CHEMOSTRATIGRAPHIC PROFILES AND CORRELATIONS .............................. 182

vii
 2. LIST OF FIGURES

Figure 1.1: (a) Chart showing the daily oil production rate of the US from 1920 to 2017 and
the historical peak, which was attained in October 1970. The contribution of shale
boom in the daily US oil production (purple color) post 2000 is also shown. (b)
Chart showing a breakdown of the daily production of the different US shale
plays and their relative contribution in the total US production. The chart shows,
as of 2017, the Bakken contributes about 23% of the daily shale play production
from the US. (Data Source for charts EIA, 2017) .......................................................... 2
Figure 2.1: Location map of the Williston Basin with subsea structure contour on the base
of the Mississippian. Elm Coulee Field is located in northeast Montana,
demarcated by the black ellipse . The limit of the Bakken Formation is shown by
a dashed line. P=location of the Parshall field area; A=location of Antelope
field.(modified from Sonnenberg and Pramudito, 2009). ............................................ 13
Figure 2.2: Late Devonian-Early Mississippian stratigraphic column of the Williston Basin
showing the Bakken Formation and the underlying Three Forks Formation and
overlying Lodgepole Formation. (modified from Webster (1984) and Sonnenberg
and Pramudito (2009)). ................................................................................................ 13
Figure 2.3: Isopach map of the Middle Bakken Member with the location of the wells studied
in this paper. The extent of the Elm Coulee Field is outlined by the green line.
The red line is the boundary of the Richland County. Location of the wells which
were studied for core description and routine core analysis (RCA) and X-ray
diffraction (XRD) derived mineralogy were examined are shown by the yellow
circles. Abbreviation of the well names shown in this map is given in Table 2.1. ...... 15
Figure 2.4: The nine different lithofacies identified in the Middle Bakken Member (MBM)
of the Bakken Formation in the Elm Coulee Field are shown in the core
photographs. Detail description of each facies is listed in Table 2.2. .......................... 18
Figure 2.5: Core description of well Coyote Putnam showing lithofacies variation in the
Middle Bakken along with the corresponding depth profiles of porosity and
saturation obtained from routine core analysis and dolomite, calcite and clay
content obtained from XRD analysis. MB-Middle Bakken, UBS-Upper Bakken
Shale. Details about the degree of dolomitization is discussed later in Section
2.5.5.............................................................................................................................. 19
Figure 2.6: The paragenetic chart is showing the different diagenetic events that the Middle
Bakken Member (MBM) has undergone in the Elm Coulee Field. ............................. 21
Figure 2.7: (a) The core photograph is showing patchy calcite cementation (white dashed
line) in the Middle Bakken facies B. CP, 10365.2ft. (b) The photomicrograph is
showing the patchy calcite (red) cement in the Middle Bakken facies B. Plane
light, BEL,10411.3ft. (c) The photomicrograph is showing micritization envelope
which developed on bioclast while calcite cement precipitated within the bioclast
in the Middle Bakken facies B. Plane light, BEL, 10411.3ft. ...................................... 22
Figure 2.8: (a) The photomicrograph is showing dolomitized matrix with unreplaced
calcareous bioclasts. Plane light, BEL,10414.2ft. (b) The photomicrograph is
showing zoned, planar euhedral dolomite rhomb in the matrix of the Middle
Bakken facies B. Zoning in dolomite rhomb is shown by the pink arrows. Plane
light, CP,10353.5ft. ...................................................................................................... 22

viii
Figure 2.9: The photomicrographs are showing different forms of pyrite in the Middle
Bakken in Elm Coulee Field. (a) Pyrite within dolomite rhomb is indicated by the
yellow arrow, whereas pyrite around the edge of the dolomite rhomb is shown by
the white arrow. Pink haze in the matrix is due to impregnation of the thin section
by red dye and indicates microporosity. Plane light, CP,10353.41ft (b)Pyrite is
replacing calcareous fossil fragment. Plane light, BEL-10414.2ft. (c) and (d)
Bioclast fragment is replaced by two generations of cement, microcrystalline
silica (white arrow) and pyrite (yellow arrow). Plane light (c) and crossed polar
(d). CP-10371.93ft. ...................................................................................................... 23
Figure 2.10: Scanning Electron Microscope (SEM) image is showing: (a) authigenic quartz
overgrowth (yellow arrow), and (b) authigenic pore bridging illite in
intercrystalline pore between the dolomite rhombs. CP4, 10350.3ft. .......................... 23
Figure 2.11: Secondary anhydrite in different forms in the Middle Bakken in the Elm Coulee
is shown in, (a) The core photograph is showing the anhydrite pseudomorphs
burrows. CP, 10354.1ft (yellow arrow). (b) The core photograph is showing the
anhydrite nodule (yellow arrow). LT, 10409.5ft. (c) and (d) The
photomicrographs are showing poikilotopic anhydrite cement in plane light and
crossed polars respectively (yellow arrow). (e) and (f) The photomicrographs are
showing poikilotopic anhydrite cement in plane light and crossed polars
respectively. Note the euhedral dolomite rhombs are floating in the anhydrite
cement (yellow arrow). ................................................................................................ 24
Figure 2.12: δ18O versus δ13C crossplot comparing the completely dolomitized samples which
were collected in this study from the Middle Bakken in Elm Coulee Field, with
the reported values in the literature. Blue and green rectangles are stable isotope
values of carbon and oxygen obtained from the Late Devonian and Early
Mississippian unaltered brachiopods respectively from Mii et al.(1999) and Van
Geldern et al. (2006). Gray shaded rectangle indicates expected stable isotope
values in dolomite which is in equilibrium with the Late Devonian sea. .................... 25
Figure 2.13: Box plot of routine core analysis (RCA) derived porosity of the Middle Bakken
in the Elm Coulee Field, showing the porosity variation observed in the field. The
median porosity of each of the facies is shown by a white line within the box
plots. Data consists of samples from all the facies from eight cores whose
locations are shown in Figure 2.3. ............................................................................... 27
Figure 2.14: The photomicrographs are showing textural features corresponding to different
dolomite content in the Middle Bakken facies B in the Elm Coulee Field. (a)
Intercrystalline pores between dolomite rhombs which are sized between10-15
µm are pointed by the yellow arrows. These pores are visible in petrographic thin
section in completely dolomitized intervals. Plane light, SJ, 9731.35ft. This
sample has RCA porosity of 7.07% and dolomite content of 50 wt.% obtained
from XRD. (b) Intercrystalline pores between dolomite rhombs which are sized
between10-15 µm are shown by the yellow arrows. These pores are visible in
petrographic thin section in completely dolomitized intervals. Plane light, CP,
10356.42ft. This sample has RCA porosity of 7.86% and dolomite content of 58
wt.% obtained from XRD. (c) Microporosity (sub-micron sized) between
dolomite rhombs in the matrix is enhanced by the red dye in petrographic thin
section. Micron sized pores are not present. Plane light, BY-10472.5ft. The
sample has RCA derived porosity of 5.98% and dolomite content of 45 wt.%
obtained from XRD. (d) Orange hue is showing the microporosity (sub-micron
sized) between dolomite rhombs in the matrix. Epifluorescence, BY-10472.5ft.

ix
 This sample has RCA derived porosity of 5.98%, and XRD derived dolomite
content of 45 wt.%. (e) Microporosity is not visible in thin section impregnated
with red dye. Plane light, CP-10367.16 This sample has RCA porosity of 3.73%
and dolomite content of 36 wt.% obtained from XRD. (f) No visible porosity in
thin section impregnated with the red epifluorescence dye. Epifluorescence, CP-
10367.16ft. Sample has RCA derived porosity of 3.73%, and XRD derived
dolomite content of 36 wt.%. ....................................................................................... 28
Figure 2.15: The Scanning Electron Microscope (SEM) photomicrographs are showing pore
architecture in the Middle Bakken facies B in the Elm Coulee Field. (a)
Intercrystalline pores between dolomite rhombs are shown by yellow arrows. CP
(2), 10356.42ft. This sample has RCA porosity of 7.86% and dolomite content of
58 wt.% obtained by XRD. (b) Yellow arrows show intercrystalline pores
between the dolomite rhombs, which is filled by pore bridging authigenic illite.
The presence of illite in between the dolomite rhombs results in a complex pore
network and a decrease in the pore throat size. CP (6), 10359.29ft. This sample
has RCA porosity of 8.85% and dolomite content of 50 wt.% obtained from XRD.
(c) Intercrystalline pore between dolomite rhombs which is shown by the yellow
line is filled by pore filling authigenic illite. CP4, 10350.3ft. The sample has RCA
derived porosity of 5.93%, and XRD derived dolomite content of 51 wt.%. (d)
Intercrystalline slot pores (yellow line) between dolomite rhombs which is filled
by pore bridging authigenic illite. CP4, 10350.3ft. This sample has RCA derived
porosity of 5% and dolomite content of 51 wt.% obtained from XRD. ....................... 30
Figure 2.16: Crossplot between X-ray diffraction (XRD) derived dolomite content, and routine
core analysis (RCA) derived total porosity. Samples are color-coded by dolomite
index (DI), which is a measure of the degree of dolomitization. Porosity increases
as dolomite content increase up to 60 wt.%. Beyond that, porosity starts
decreasing as dolomite content increases more than 60 wt.%. .................................... 31
Figure 2.17: Pore throat size distribution of the Middle Bakken samples from the Elm Coulee
Field determined by Mercury Intrusion Porosimetry (MIP) is shown. Pore throat
size distribution is color-coded by X-ray diffraction (XRD) dolomite content.
Porosity and mineralogy details of the samples are given in Table 2.3. ...................... 31
Figure 2.18: (a) Subhedral, partially coalesced micritic particles of precursor limestone,
corresponding to the Stage I of porosity evolution. Intercrystalline sub-micron
sized pores between the micrite particles. BEL, 10402ft, Dol-13 wt.%, Calc-58
wt.%, porosity-3.64%. (b) Remnant microcrystalline calcite from the precursor
micritic lime mud is present in between the euhedral dolomite rhombs in the
partially dolomitized Middle Bakken facies B (stage II of porosity evolution).
Dolomite rhombs are 15-20 µm in size. Note that micritic calcite is in direct
contact with the dolomite rhombs. BY-13, 10457.6ft, Dol-36 wt.%, Calc-24
wt.%, porosity-2.39%. (c) Subhedral to anhedral fully coalescent remnant
microcrystalline calcite from precursor lime mud is in direct contact with the 20
µm sized dolomite rhombs in the partially dolomitized Middle Bakken facies B
(stage II of porosity evolution). (d) Intercrystalline pores between the euhedral
dolomite rhombs, in the completely dolomitized interval of the Middle Bakken
corresponding to the stage III of porosity evolution, is shown by yellow arrows.
Dolomite rhombs are around 30-40 µm in size. CP (2), 10356.42ft. The sample
has RCA derived porosity of 7.86%, and XRD derived dolomite content of 58
wt.%. (e) Euhedral dolomite rhombs from the overdolomitized interval (stage IV
of porosity evolution) forming interlocking texture with compromised crystal

x
 boundaries between dolomite rhombs is shown by the black arrow. Dolomite
rhombs are more than 100 µm in size. BY-14, 10457.6ft, Dol-66 wt.%, porosity-
4.25%. (f) Interlocking texture with compromised crystal boundaries (pink arrow)
between euhedral dolomite rhombs in the overdolomitized interval (stage IV of
porosity evolution) of the Middle Bakken is shown by black arrows. Dolomite
rhombs are more than 100 µm in size. CP-3, 10353.25ft, Dol-63 wt.%, porosity-
4.4%. ............................................................................................................................ 35
Figure 2.19: Schematic diagram illustrating an increase in the size of dolomite rhombs results
in an increase in pore size. ........................................................................................... 39
Figure 2.20: Conceptual model illustrating porosity evolution both spatially and temporally
due to seepage reflux dolomitization in a ramp setting. As seepage reflux
continues from time T1 through T2 continues, limestones are progressively
dolomitized, with the formation of dolomite cement near the source of the brine
and anhydrite cement at the distal end, which in turn results in variation of
porosity. (Jones and Xiao, 2005).................................................................................. 40
Figure 2.21: (a) At time T1, the precursor lime mud got deposited, and the entire sediment
column has an overall high porosity. (b) At time T2, early patchy calcite
cementation results in porosity loss at depths between d0 to d1. (c) At time T3,
seepage reflux of mesohaline brine has started. Sediment column from d3 to d2 is
partially dolomitized, with the degree of dolomitization decreasing down the
depth. No change in total porosity is observed between lime mud and partially
dolomitized limestone due to dolomitization. However, pore space and pore size
are modified as intercrystalline pores start to form between dolomite rhombs.
Moreover, the formation of anhydrite cement may result in minor loss of porosity
near depth d1. (d) At time T4, seepage reflux of mesohaline brine continues such
that the partially dolomitized lime mud between d3 to d2 is completely
dolomitized with no change in total porosity. While lime mud between d2 to d1 is
partially dolomitized with the degree of dolomitization decreasing down the
depth such that below depth d0, precursor lime mud remains unaltered. There is
no change in total porosity due to dolomitization. However, pore space and pore
size are modified as intercrystalline pores forms between dolomite rhombs.
Moreover, the formation of anhydrite cement may result in minor loss of porosity
between depth d1 to d0. (e) At time T5, seepage reflux of mesohaline brine
continues such that the completely dolomitized interval between d3 to d2 is
overdolomitized with a loss in porosity due to the formation of dolomite cement.
While partially dolomitized lime mud between d2 to d1 is completely dolomitized.
Below depth d1, the precursor lime mud is partially dolomitized with the degree
of dolomitization decreasing down the depth such that below depth d0, precursor
lime mud remains unaltered. There is no change in total porosity due to
dolomitization. However, pore space and pore size are modified as
intercrystalline pores forms between dolomite rhombs. (f) At time T 5, after
cessation of the seepage reflux dolomitization, mechanical compaction of the
sediment column occurs due to burial. This results in differential compaction of
the sediment column such that the precursor lime mud is more compacted and
therefore, loses more porosity in comparison to the dolostones. ................................. 42
Figure 3.1: (a) Map showing the distribution of the Bakken Formation in the Williston Basin
and its equivalent Exshaw Formation and Lower Banff Formation in the Western
Canada Sedimentary Basin (modified from Smith and Bustin, 2000). (b)
Locations of the wells used in this study are shown in the isopach map of the

xi
 Upper Bakken Shale in the U.S. portion of the Williston Basin (source: Stephen
Sonnenberg, personal communication). Abbreviations of the names of the wells
used in this study are listed in Table 3.1. ..................................................................... 53
Figure 3.2: Generalized stratigraphic column of the Williston Basin with an expanded section
showing the stratigraphic details along with the conodont zones of the Late-
Devonian-Early Mississippian age Bakken Formation (Sandberg et al., 1988;
Karma, 1991; Hartel et al., 2012; Sonnenberg, 2015).................................................. 54
Figure 3.3: (a) Core photograph illustrating massive radiolarian-dominated siliceous
mudstone (F1a). The characteristic ptygmatic pyrite-filled, sub-vertical fractures
in facies F1a are shown by white arrows. Well: Rasmussen 1-21-6H, 10211ft. (b)
Core photograph illustrating finely laminated radiolarian and silt-bearing
mudstone (F1b). Well: Rasmussen 1-21-6H, 10214ft.................................................. 62
Figure 3.4: (a) Scanned petrographic thin section image showing ~14mm thick bed of
massive radiolarian dominated mudstone (F1a). The sharp contact at the base and
top of the facies F1a are indicated by the white arrow; Well: LTE, 10386 ft. (b)
Optical photomicrograph of F1a showing the typical fabric of the massive
radiolarian dominated siliceous mudstone (F1a). The high content of radiolarians,
which are shown by the red arrow is evident from this photomicrograph. Wispy
organic matter surrounding the radiolarians is shown by the yellow arrow. (c)
Thin section photomicrograph of facies F1a showing radiolarian tests, which was
replaced by chert (green arrow), sphalerite (yellow arrow) and pyrite (red arrow).
Well: RT, 10663.60 ft. (d) Optical photomicrograph of facies F1a showing is
radiolarians, which are replaced by calcite (green arrow) and pyrite (red arrow).
As evident, it is difficult to distinguish the shape of the individual radiolarians as
they are mostly dissolved out. Well: RT, 10661.9 ft.................................................... 63
Figure 3.5: (a) Optical photomicrograph showing a vertical ptygmatic fracture (red arrow) in
massive radiolarian dominated mudstone (F1a) Well: RT, 10663.60ft. (b) Optical
photomicrograph showing the vertical fracture (red arrow) in Figure 3.5a at
higher magnification. Two generations of cement within the fracture, sphalerite
(red arrow) followed by pyrite (yellow arrow) is observed. Well: RT, 10663.6ft. ...... 64
Figure 3.6: (a) Scanned petrographic thin section image of laminated radiolarian and silt-
bearing mudstone (F-1b). The double-headed yellow arrows show sub-mm scale
laminae of F1b. Red arrows showing sphalerite-filled vertical fractures in the
middle of the image Well: LT, 10390.1ft. (b) An optical photomicrograph of
laminated radiolarian and silt-bearing mudstone (F-1b) showing discontinuous
laminae of radiolarians (red arrows). Radiolarians (yellow arrows) along with silt
grains are observed to be randomly distributed in an organic matter and clay-rich
matrix in the lower part of the image. Well: LT, 10390.1ft. ........................................ 65
Figure 3.7: (a) Core photograph showing the massive to finely laminated silt-bearing
mudstone (F2). Discontinuous mm-size pyrite lenses (shown by the yellow
arrow) is the most distinct macro-scale characteristics of F2. Well: R, 1265ft. (b)
Optical photomicrograph of massive to finely laminated silt-bearing mudstone
(F2) showing the presence and fecal pellets (p) and silt grains which are randomly
dispersed in the matrix. Well LT, 10384.6ft, TOC: 14.3 wt.%. (c) Optical
photomicrograph showing massive to finely laminated silt-bearing mudstone
(F2a), which have characteristically high matrix content and proportionately less
fecal pellets (<5%) compared to massive to finely laminated fecal pellet and silt-
bearing mudstone (F2b). Well: BN, 10352.1ft. (d) Optical photomicrograph

xii
 showing massive to finely laminated fecal pellet and silt-bearing mudstone (F2b),
which have proportionately higher amounts of fecal pellets (p) (>5%) compared
to massive to finely laminated silt-bearing mudstone (F2a). Well: 4-11-7-13W2,
1776.25ft, TOC: 28.6 wt.%. ......................................................................................... 66
Figure 3.8: (a) FE-SEM image of massive to finely laminated silt-bearing mudstone (F2)
showing the presence of pyrite framboids (pyr fram), which are less than 10µm
in size and are randomly dispersed in clay and organic matter-rich matrix. Well:
RT, 10668ft, TOC:14.7 wt.%. (b) Optical photomicrograph showing the
occasional presence of radiolarians (rad) in massive to finely laminated silt-
bearing mudstone (F2). Well: RT, 10665.5ft, TOC: 12 wt.%. ..................................... 67
Figure 3.9: (a) Optical photomicrograph from a parallel to bedding thin-section from massive
to finely laminated silt-bearing mudstone (F2) showing the cross-section of type
1 fecal pellet, which is cylindrical in shape, clay-rich and organic poor. Well: LT,
10388.1ft, TOC: 11.2 wt.%. (b) Optical photomicrograph from a parallel to
bedding thin-section from massive to finely laminated silt-bearing mudstone (F2)
showing the cross-section of type 2 fecal pellet, which is circular-shaped and silt-
rich. Well: LT, 10388.1ft, TOC: 11.2 wt.%. (c) FESEM image showing the shape
of type 1 fecal pellet, which is cylindrical in shape and 200µm in length. Well:
RT, 10665.6ft, TOC: 10.5 wt.%. (d) The SEM-EDAX image showing the near
absence of organic matter (red) within the type 1 fecal pellet, while the matrix is
rich in organic matter. Well: RT, 10665.6ft, TOC:10.5 wt.%...................................... 68
Figure 3.10: (a) Core photograph showing continuous and discontinuous lamination (yellow
arrow) of sand and clay-bearing, silt-rich mudstone (F3). (b) Scanned image of a
petrographic thin section showing planar silt laminations (yellow arrow) in the
sand and clay-bearing, silt-rich mudstone (F3) well-LT, depth-10389ft (c) Optical
photomicrograph showing framework components of very-fine sand to coarse silt
grains in a clay and organic-rich matrix in facies F3. (d) An optical micrograph
showing massive to finely graded beds of facies F3. Presence of phosphatic
fragments is indicated by the yellow arrows. Well: LT, 10387.1ft. (e) The optical
photomicrograph showing starved ripples and continuous planar silt lamina.
Well: LT, 10389.8ft...................................................................................................... 70
Figure 3.11: (a) Optical photomicrograph showing lenticular silt beds with a sharp base and
overlain by a clay-rich lamina. The sample is from well LT, depth 10389ft. (b)
The optical photomicrograph is showing horizontal Planolites burrow indicated
by yellow arrow in the clay-rich lamina overlying the silt lamina. Well: RT,
10670ft. ........................................................................................................................ 71
Figure 3.12: (a) Core photograph of macrofossil-bearing silt-rich mudstone (F4) showing
horizontally aligned thin-shelled brachiopods (brac), which is a diagnostic
characteristic of this facies. Well: RT, 10654ft. (b) Core photograph of
macrofossil-bearing silt-rich mudstone (F4) showing thin-shelled brachiopods in
facies F4. Well: L, 10937.2ft. (c) The optical photomicrograph showing
articulated brachiopods (brac) in facies F4. Brachiopods are partial to completely
pyritized. Well: RT, 10654.5ft. (d) The optical photomicrograph showing
articulated brachiopods (brac) in facies F4. Brachiopods are replaced by silica
and pyrite. Well: RT, 10653.5ft. .................................................................................. 72
Figure 3.13: (a) and (b) Optical photomicrograph in plane polars and crossed polars showing
the presence of abundant silt content and homogeneous fabric in facies F4.
Homogeneous fabric in facies F4 is probably a result of meiofaunal burrowing.

xiii
 Abundant detrital dolomite (dd), authigenic dolomite and agglutinated benthic
foram(af) in this facies are observed at the lower right corner of the image. Well:
RT, 10653.5ft TOC: 2.78 wt.%. (c) and (d) Optical photomicrograph in plane
polar and crossed polars, showing agglutinated benthic forams (af) in the facies
F4 at a higher magnification. Presence of quartz silt in the wall of the benthic
foram can be seen in the wall of benthic foram as pointed by the yellow arrow.
Well-RT, 10650ft, TOC: 7.34 wt.%. ............................................................................ 73
Figure 3.14: (a) Optical photomicrograph is showing few silt grains thick discontinuous planar
laminae with downlapping surface (shown by the yellow arrow) in macrofossil-
bearing silt-rich mudstone (F4). Well: RT, 10354.5ft (b) Core photograph
showing the presence of macroscale Planolites burrows (pl) in macrofossil-
bearing silt-rich mudstone (F4). Well: L, 10934ft. ...................................................... 74
Figure 3.15: (a) Core photograph showing macro-scale Planolites burrows (b) in burrow-
mottled silt-bearing mudstone (F5) at the contact of UBS with the overlying
Lodgepole Formation. Well: LT, 10383.2ft. (b) A core photograph is showing
macro-scale Planolites burrows (b) in burrow-mottled silt-bearing mudstone (F5)
at the contact of UBS with the overlying Lodgepole Formation. Well: RT,
10648ft. (c) Scanned image of petrographic thin section showing Planolites
burrows (b) in facies F5. Well: LT, 10383.2ft. (d) Optical photomicrograph of
facies F5 is showing that the burrow (b) filled by constituents, which are similar
to the matrix, and is devoid of organic matter. Well: LT, 10383.2ft. .......................... 75
Figure 3.16: (a) Core photograph showing mm-thick bed of phosphate clast and fossil-rich
mudstone (F6). Well: R, 11026.6ft. (b) Core photograph showing cm-thick bed
of phosphate clast and fossil-rich mudstone (F6). Well: BN, 10358.4ft. (c) Core
photograph is showing cm-thick bed of phosphate clast and fossil-rich mudstone
(F6) at the contact of UBS with the Middle Baken. Well: R, 11044.6ft. (d) Optical
photomicrograph showing phosphate clasts, and sponge spicules (shown by red
arrow) oriented randomly in phosphate clast and fossil-rich mudstone (F6). The
base of this facies shows microscouring. Well: BN, 10352ft. .................................... 76
Figure 3.17: Four groups of element clusters shown on the crossplot between the two principal
components determined from Principal Component Analysis (PCA) in
XLSTATTM of twenty elements. The PCA was performed on around 400 UBS
sampled from sixteen cores. ......................................................................................... 77
Figure 3.18: (a) Crossplot of Al vs. Ti showing a strong positive covariance between the two
elements, which indicates that Al has a detrital origin in UBS. (b) Crossplot of K
vs. Ti showing strong positive covariance between the two elements indicating K
has a detrital origin. Each crossplot includes all the samples from the sixteen wells
for, which XRF data were collected in this study. All the data points are color-
coded by the three different chemostratigraphic units of UBS which are described
in Section 3.4.5. All the three crossplots form two clusters, which correspond to
the two main chemostratigraphic units, namely Unit-1 and Unit-2. ............................ 79
Figure 3.19: (a) Cross-plot of Al vs. K showing a strong covariance between the two elements,
which indicates that most of the Al and K are associated with clay minerals in
UBS. (b) Crossplot between Ti vs. Si showing a negative correlation, which
indicates that f Si has both non-detrital origin biogenic/authigenic origin and
detrital origin. All the data points are color-coded by the three different
chemostratigraphic units of UBS which are described in Section 3.4.5. All the

xiv
 three crossplots form two clusters, which correspond to the two main
chemostratigraphic units, namely Unit-1 and Unit-2. .................................................. 79
Figure 3.20: Box and whisker plot showing the variation in the absolute concentration for the
three redox-sensitive trace elements: Mo, U, and V in UBS. Each plot comprises
of all samples of UBS from all the sixteen wells for which XRF data was collected
in this study. ................................................................................................................. 80
Figure 3.21: (a) Crossplot of Mo vs. U showing a strong positive covariance between the two
elements. (b) Cross-plot between Fe and S showing a positive covariance which
indicates that most of the Fe and S are associated with pyrite. DOPT (Degree of
Pyritization) lines in the crossplot are based on Algeo and Maynard (2004). All
the samples are color-coded by the three different chemostratigraphic units of
UBS, which are described in Section 3.4.5. ................................................................. 80
Figure 3.22: (a) Cross-plot of Mo vs. V showing a weak covariance between the two elements.
The data points from three different clusters which are indicated in by
encirclements. (b) Cross-plot of U vs. V showing a weak correlation between the
two elements. The data points from three different clusters are indicated by
encirclements. All the samples are color-coded by the three different
chemostratigraphic units of UBS, which are described in Section 3.4.5. .................... 81
Figure 3.23: Depth profiles comparing the XRF data of major elements and XRD mineralogy
of UBS in (a) Well LT, and (b) Well BF. .................................................................... 82
Figure 3.24: Core description and chemostratigraphic profile with elemental trends of
biogenic, detrital, carbonate and redox-proxying elements from Rolf 1-20H core
showing the chemostratigraphic sub-unit 1a, sub-unit 1b and Unit-2 of UBS.
From left to right: Si-silicon; Al-aluminum; K-potassium; Ti-Titanium; Zr-
zirconium; Ca-calcium; Mo-molybdenum; U-uranium; V-vanadium. MFS-
Maximum Flooding Surface. ....................................................................................... 84
Figure 3.25: Core description and chemostratigraphic profile with elemental trends of
biogenic, detrital, carbonate and redox-proxying elements from Roberts Trust 1-
13H showing the chemostratigraphic sub-unit 1a, sub-unit 1b and Unit-2 of UBS.
From left to right: Si-silicon; Al-aluminum; K-potassium; Ti-titanium; Zr-
zirconium; Ca-calcium; Mo-molybdenum; U-uranium; V-vanadium; MFS-
Maximum Flooding Surface. ....................................................................................... 85
Figure 3.26: Core description and chemostratigraphic profile with elemental trends of
biogenic, detrital, carbonate and redox-proxying elements from core Graham
USA 1-15 showing the showing the chemostratigraphic sub-unit 1a, sub-unit 1b
and Unit-2 of UBS. From left to right: Si-silicon; Al-aluminum; K-potassium; Ti-
titanium; Zr-zirconium; Ca-calcium; Mo-molybdenum; U-uranium; V-vanadium.
MFS-Maximum Flooding Surface. .............................................................................. 86
Figure 3.27: NE-SW chemostratigraphic correlation between wells RR Lonetree Edna, Round
Prarie 1-17H, Rolf 1-20H, Sidonia 1-06H and Jorgensen 1-15H. ............................... 89
Figure 3.28: NW-SE chemostratigraphic correlation between wells Rasmussen 1-21-16H,
Charlotte 1-22H, Roberts Trust 1-13H, and Miller 34X-9. .......................................... 90
Figure 3.29: (a) Isochore map of the UBS showing north-northwest to south-southeast trend
of maximum thickness in the basin depocenter. (b) Isochore map for sub-unit 1a
showing a north-northwest to south-southeast trend of maximum thickness in the
basin depocenter. Maximum thickness for sub-unit 1a is encountered in well C

xv
 (6.5ft). (c) Isochore map of chemostratigraphic sub-unit 1b showing a north-
northwest to south-southeast trend of maximum thickness in the basin depocenter.
Maximum thickness for sub-unit 1b is encountered in well L (9ft). (d) Isochore
map of chemostratigraphic Unit-2 showing the development of maximum
thickness around wells L and RT in the southern part of the basin depocenter.
Unit-2 loses its thickness drastically along the southern basin margin. Maximum
thickness for Unit-2 is encountered in well RT (9.2ft). An inset map in each figure
showing the isopach map of the entire UBS in US-portion of the Williston Basin
and the shaded box in the inset map represents the study area. Locations of wells
are shown by black dots. Abbreviations of well names used in the maps are listed
in Table 3.1. ................................................................................................................. 91
Figure 3.30: Correlation is showing the lateral and vertical facies distribution in the three
different chemostratigraphic units. The boundaries of the three units marked in
this correlation are picked from the chemostratigraphic profiles. ................................ 94
Figure 3.31: Correlation is showing the lateral and vertical facies distribution in the three
different chemostratigraphic units. The boundaries of the three units marked in
this correlation are picked from the chemostratigraphic profiles of the wells. ............ 95
Figure 3.32: Vertical profile of gamma ray (GR) log and SRA derived total organic carbon
(TOC) for UBS for wells: RT, BF, and S showing the presence of high GR peak
near the base and top of UBS and an upward decreasing trend in the middle of
UBS interval in all the three wells. In well RT, TOC profile mimics GR trend,
whereas in the other two wells variation in TOC is insignificant and TOC profile
does not follow GR. ..................................................................................................... 96
Figure 3.33: Histograms showing the statistical distribution of Mo concentration in various
chemostratigraphic units of UBS. a) For sub-unit 1a, more than 90% of the
samples have Mo concentration more than 100ppm; b) For sub-unit 1b, more than
80% of the samples have Mo concentration more than 100ppm; b) For Unit-2,
more than 60% of the samples have Mo concentration more than 100ppm. ............ 101
Figure 3.34: Ternary plot representing 5*Al2O3-SiO2-2*CaO relationship (after Brumsack,
1989) of the chemostratigraphic sub-unit 1a, 1b, and Unit-2 of UBS. The three
axes of (5*Al2O3), SiO2 and (2*CaO) represents clays, quartz and/or biogenic
silica, and total carbonates respectively. The black data points represent XRD
mineralogy from five wells, shape of the data points represents the three different
chemostratigraphic units. The ternary plot includes all the samples from the
sixteen wells for which XRF data has been collected in this study. All the data
points are color-coded by the three different chemostratigraphic units of UBS.
The data points form mainly two clusters, which correspond to the two main
chemostratigraphic units: Unit-1 and Unit-2. Sub-unit 1b predominantly clusters
below the 50% clay line, whereas Unit-2 plots above the 50% clay line................... 108
Figure 4.1: Location of well Robert Trust 1-13H (RT) is shown by the red star in the isopach
map of the Upper Bakken Shale (modified from Sonnenberg et al., 2017). .............. 121
Figure 4.2: Schematic S-Fe-C ternary diagram showing the different degrees of pyritization
(DOP) regions for suboxic, anoxic and euxinic conditions (modified from
Rimmer et al., 2004)................................................................................................... 126
Figure 4.3: Integrated profile of UBS in core Roberts Trust 1-13H (RT) showing different
chemostratigraphic units of UBS, which were distinguished based on the vertical
distribution of lithofacies, detrital and biogenic proxying elements of Si, Al, K,

xvi
 Ti, Zr/Rb ratio, Ca and redox-proxying elements of Mo and U. Characteristics of
each chemostratigraphic sub-unit are summarized in Table 4.1. The profile of
TOC, stable isotope of δ15Norg, which is a proxy for the paleoproductivity, and
stable isotope of δ13Corg, which acts as the proxy for the OM type, are also shown.
TOCSRA represents TOC measured from source rock analysis (SRA) and TOCorg
represents original TOC calculated from TOCSRA. The basis for identifying the
Maximum flooding surface (MFS) was described in Section 3.5.5. .......................... 129
Figure 4.4: Samples from Well RT plotted on the modified Van Krevelen diagram showing
sub-unit 1a, 1b and 2b predominantly consist of type II-III organic matter, while
sub-unit 2a has a higher proportion of type III organic matter. ................................. 131
Figure 4.5: Box and whisker plot showing the stable isotope of carbon (δ13Corg) results for
the different chemostratigraphic UBS units in Well RT. Mean values of δ 13Corg
for each of the chemostratigraphic sub-units are labeled to the right of each of the
boxes. Average values for δ13Corg terrestrial-type III (-25‰ to -26 ‰) and marine-
type II (-30 ‰) OM are also shown from Chen and Sharma (2016) by the green
rectangle and brown line respectively. ....................................................................... 132
Figure 4.6: Cross plot between Ti and Al show a positive linear covariance for Unit-1, which
indicates that Al has a detrital origin. (b) Crossplot between Ti and K show a
positive linear covariance for Unit-1, which indicates that K has a detrital origin,
and for this unit, K can be used as a proxy for detrital illite clay. (c) Crossplot
between Al and ������ show weak positive covarriance, which indicates that
influx of clay did not result in dilution of organic matter as Al is a proxy for
detrital clay minerals in UBS (Section 3.4.2, Chapter-3). (d) Crossplot between
K and ������ show no covarricance, which confirms that influx of clay did not
result in dilution of organic matter as K is a proxy for detrital clay minerals in
UBS (Section 3.4.2, Chapter-3). (e) Crossplot between Ti and Al show a positive
linear covariance for Unit-2, which indicates that Al has a detrital origin in this
unit also. (f) Crossplot between Ti and K show a positive linear covariance for
Unit-2, which indicates that K has a detrital origin, and for this unit, K can be
used as a proxy for dertrital illite clay. (g) Crossplot between Al and ������ for
Unit-2 show a positive linear covariance, which indicates that influx of clay
helped in organic matter preservation in Unit-2 as Al is used as proxies for detrital
clay minerals (h) Crossplot between K and ������ for Unit-2 show a positive
linear covariance, which confirms that influx of clay helped in organic matter
preservation in Unit-2. ............................................................................................... 133
Figure 4.7: (a) Crossplot between Ca and ������ for Unit-1 show no correlation indicating
influx of carbonates did not affect organic matter preservation. (b) Crossplot
between Ca and ������ for Unit-2 showing a negative linear correlation, which
indicatcates influx of carbonates especially detrital dolomites resulted in dilution
of organic matter. ....................................................................................................... 134
Figure 4.8: For Unit-1, crossplots between (a) Si and Al, and (b) Si and Ti, showing no
covariance, which indicates in Unit-1 has a considerable proportion of Si is
associated with the radiolarians in siliceous mudstone (F1). It also indicates that
biogenic silica had a dilution effect on the detrital clay. (c) Crossplot between Si
and ������ for Unit-1 show a negative correlation indicating influx of biogenic
Si had resulted in the dilution of organic matter. For Unit-2, crossplot between
(d) Si and Al (e) Si and Ti show positive correlation indicating Si in Unit-2 has
detrital origin. (f) Crossplot between Si and ������ for Unit-2 showing a

xvii
 negative correlation, which indicates influx of detrital Si associated with silt-sized
sediment had resulted in the dilution of organic matter. ............................................ 135
Figure 4.9: (a) Crossplot between total iron (FeT) and total sulfur (ST) for different
chemostratigraphic sub-units of UBS is shown. ST/FeT stochiometric ratio of 1.15
represents pyrite shown by the black solid line in the plot, and the equivalent
degree of pyritization (DOP) is equal to 1 for this line, which indicates the euxinic
condition. The green broken line represents regression fit for Unit-1and black
broken line represents the regression fit for Unit-2. Unit-1 plots close to DOP=1
line indicating that most of the iron in Unit-1 is associated with pyrite and this
unit was deposited in euxinic condition. Whereas Unit-2 plots below the DOP=1,
which indicates that supply of iron was more compared to sulfur and pyrite
formation was sulfur limited. This indicated Unit-2 was deposited at
comparatively less reducing conditions than Unit-1. (b) Crossplot between
������ and total sulfur (ST) for the different chemostratigraphic units of UBS is
shown. Solid line represents S and C trend for oxygeneted normal marine
conditions during organic matter deposited. Broken lines represent linear
regression data for Unit-1 and Unit-2. The non-zero intercept of the linear
regression lines for both Unit-1 and Unit-2 in the crossplot indicates that their
deposition has taken place in reducing condition and not in oxygeneted condition.
.................................................................................................................................... 136
Figure 4.10: Fe-S-TOCorg ternary diagram for UBS samples from well RT showing sub-unit
1a and 1b mostly plot along DOP~0.96, which indicates a strongly euxinic
condition for these sub-units. Sub-unit 2a plots along DOP~0.56 suggesting
suboxic condition for this sub-unit, while sub-unit 2b plots between DOP of 0.56-
0.75, suggesting for 2b the paleoredox condition fluctuated from suboxic to
anoxic with increasing intensity of anoxia compared to that in sub-unit 2a. ............. 137
Figure 4.11: Crossplot between (a) ������ vs Mo; and (b) ������ vs U showing three
distinct covariation patterns. (c) Chart showing the expected pattern of relative
enrichment of Mo and U at suboxic, anoxic and euxinic conditions (modified
from Tribovillard et al., 2006), which suggest in suboxic condition, ������ <
5 wt.%; Mo has no enrichment while U shows limited enrichment. In anoxic
condition, 5 < ������ < .5 wt.%, both Mo and U shows moderate
enrichment and has a good linear covariance pattern with ������. In euxinic
condition, ������ > .5 wt.%, concentration of Mo and U increases
exponentially and shows poor correlation with ������........................................... 138
Figure 4.12: Crossplot between U Enrichment factors (U-EF) vs. Mo Enrichment factors (Mo-
EF) showing three clusters of the UBS samples from core RT. The broken lines
represent 0.3, 1 and 3 times of the average Mo/U of present-day seawater (SW).
The general patterns of U-EF vs. Mo-EF variation in modern marine
environments are represented by (i) gray field for unrestricted marine trend for
eastern tropical Pacific Ocean, (ii) green field for semi-restricted Cariaco Basin,
where particulate shuttle function for Mo enrichment within the water column
and (iii) black solid line for strongly restricted Black Sea (adopted from
Tribovillard et al., 2012). The yellow field represents areas, where the presence
of sulfurized organic matter leads to very high enrichment of Mo in strongly
euxinic conditions (Tribovillard et al., 2012). Sub-unit 1a and 1b plots in the zone
of sulfurized organic matter (yellow field), whereas sub-unit 2b and part of sub-
unit 2a and plots in the anoxic to the euxinic field (gray area). Finally, rest of the

xviii
 samples from sub-unit 2a plot in the sub-oxic to the anoxic field in the gray area.
.................................................................................................................................... 141
Figure 4.13: (a) Mo-TOCorg crossplot for UBS in Well RT showing slope (m) for the
regression line for chemostratigraphic sub-unit 1a, 1b and 2b are 21.1, 23.5 and
19.3 respectively. Value of m for these UBS sub-units is close to that of the semi-
restricted Cariaco Basin, which m value close to 25. (b) m versus deepwater
renewal times for modern anoxic silled basins showing an average m of 21 for
UBS suggests the deepwater renewal in the Williston Basin during the deposition
of UBS was around 25 years (modified from Algeo and Lyons, 2006). .................... 142
Figure 4.14: (a) Box and whisker plot showing the results for the stable isotope of nitrogen
(δ 5Norg for the chemostratigraphic units of UBS in well RT. The mean values
of δ15Norg for each chemostratigraphic units are also labeled. (b) TOCorg vs
δ 5Norg crossplot showing that sub-unit 1a and 1b have high TOC and lighter
values of δ 5Norg, whereas sub-units 2a has comparatively lower TOC and
heavier values of δ 5Norg. Sub-unit 2b has mixed values of δ 5Norg. This
crossplot indicates paleoproductivity during the deposition of Unit-1 was
comparatively higher than that during the deposition of sub-unit 2a. The
paleoproductivity increased marginally again during the deposition of sub-unit
2b................................................................................................................................ 143
Figure 4.15: Schematic diagram showing the prevailing depositional setting of the Williston
Basin during the deposition of UBS. The basin was under the influence of the
east-west trade wind as shown by the yellow arrow due to its location near the
equator. The Williston Basin was semi-restricted in nature due to the presence of
the Central Montana Uplift, which was situated at the south-west of the basin.
Stratified water column developed due to the formation of a thermocline, which
stabilized the chemocline, which in turn developed as a result of degradation of
organic matter. The chemocline created a stagnant euxinic bottom water
condition during the deposition of Unit-1 of UBS. However, when the depth of
the thermocline is lowered due to fall in the sea level, the seasonal fluctuation of
the thermocline disturbed the chemocline. This, in turn, decreased the reducing
condition at the sediment-water interface during the deposition of Unit-2. The
primary source of the nutrients was through regeneration of the biolimiting
nutrients N and P in euxinic condition. ...................................................................... 146
Figure 4.16: (i) Stratified water column shown developed during the deposition of sub-unit 1a
due to the formation of a thermocline. Accumulation of OM continued
incessantly, which exhausted the oxygen content in the bottom water and thereby
resulted in the development of a stagnant euxinic condition below the
chemocline. The biolimiting nutrients P and N were preferentially released in a
euxinic condition, which resulted in high surface water productivity. Seasonal
fluctuation of the depth of the thermocline also resulted in fluctuation of the
chemocline which aided in releasing the nutrients back to the surface water. (ii)
The relative rise in sea level during the deposition of sub-unit 1b increased the
height of the oxygenated surface water column which decreased the export
productivity since more OM were degraded while settling down through the
water column. Primary productivity in the surface water was high. Bottom water
conditions were still euxinic which extended to the photic zone intermittently.
(iii) The sea level drop resulted in disappearance of radiolarians, and the depth of
the thermocline dropped due to the fall in relative sea level during the deposition
of sub-unit 2a. So, the thermocline became closer to the chemocline. Seasonal

xix
 fluctuation of the thermocline destabilized the chemocline due to the partial
mixing of the water column. This resulted in the formation of sub-oxic to anoxic
condition along the sediment-water interface. Paleoproductivity decreased as the
biolimiting nutrients were less efficiently regenerated in sub-oxic condition. As
more land area got exposed due to relative sea level fall, it was covered by
vegetation which became the source of more terrestrial OM during the deposition
of sub-unit 2a. (iv) During the deposition of sub-unit 2b, there was a minor
increase in the depth of the thermocline due to a slight increase in relative sea
level. As a result, the chemocline stabilized, and conditions were more reducing
which enhanced the regeneration of the biolimiting nutrients (P and N).
Therefore, the surface water productivity was higher compared to that of sub-unit
2b................................................................................................................................ 150
Figure C.1: Photograph showing two of the thirteen UBS-matrix specific sample standard
cups, which were used in this study for XRF measurement QA-QC purpose. .......... 173
Figure C.2: Optimal detection time for the Main, Low, High and Light filters for UBS is
shown schematically. The effect of an increase in the detection duration on the
measurement of absolute concentration and the error ( �) also shown..................... 174
Figure C.3: The change in elemental concentration measurement and relative error for
different excitation filters sequences (EFS) for Si, Al, K, Mo, U, and V is shown.
Optimum time and EFS is characterized by stabilization of elemental
concentration readings and relative error values. For light elements Si and Al, the
optimum DT for the UBS samples with Niton handheld XRF were 180s and that
for other elements was 120s. ...................................................................................... 175
Figure C.4: Niton-ICP/Element Analyzer cross plots for eighteen selected elements showing
the relationships between Niton measured elemental concentrations and
laboratory-grade measurements made by ICP-MS/ICP-OES and element
analyzer. Best-fit linear regression lines are shown with associated R2 values for
each plot. For elements Al, Si, Ca, Fe, S, Mo, and U have a zero y-axis intercept.
.................................................................................................................................... 176
Figure C.5: Profile of the representative suite of major and trace elements comparing Niton
XRF and ICP-MS measurements. Visual identification of the elemental trend
between the Niton XRF and ICP-MS measurement show consistency. .................... 177
Figure C.6: Depth profiles of concentration show the repeatability test results for trace and
major elements for eleven UBS samples (represented by sample depth) from well
RT for each test runs. The error bars indicate the variability in elemental
concentration measurement is for five test run for a sample of given depth. The
red line connects the average of the five runs, whereas the blue line shows the
spread of the measurement. Repeatability of the XRF measurement using Niton
handheld XRF in TAG mode is confirmed by consistency in both major
inflections and concentration changes for five test-run.............................................. 180
Figure C.7: Repeatability of the XRF element concentration measurements for the eleven
UBS samples from core RT is shown in the heat-map, in which the numbers
represent the variation σμ The color-scale and the bar show the relative degree of
repeatability for element concentration measurement by XRF. Cooler color
(green) and lower percentages indicate better repeatability. Mg (Atomic No.-12),
which was not used in this study, in general, shows poor repeatability because
hand-held XRF and TAG mode used in this study is not suitable for detecting
elements lighter than Al (Atomic No.-13). ................................................................ 181

xx
 3. LIST OF TABLES

Table 2.1: Name and location of the wells from the Elm Coulee Field from which cores, thin
sections, routine core analysis (RCA) and X-ray diffraction (XRD) mineralogy
are used in this study. ................................................................................................... 14
Table 2.2: Sedimentary and ichnologic characteristics of the sedimentary facies identified in
the Middle Bakken in the Elm Coulee Field. ............................................................... 17
Table 2.3: Measured porosity from routine core analysis and dolomite, calcite, clay content
derived from X-ray diffraction and dolomite index (DI) of the samples analyzed
for mercury intrusion porosimetry (MIP) in Figure 2.16. ............................................ 32
Table 3.1: Names, abbreviations, APIs, and locations of the wells used in this study are
listed. The available dataset used in this study are also listed. Description of
abbreviations and symbol used in the table are as follows, Core Desc.: Core
Description, Thin Sec.: Thin-section, ✓: Acquired in this study, ✕: not available,
✓#: Available from external sources.1-USGS, 2-Fishman et al. (2015), 3-(Jin,
2014), 4-(Kocman, 2014), and 5-NDGS. Sample for well 4-11-7-13W2 was
available from Hui Jin (personal communication) ....................................................... 56
Table 3.2: Characteristic differences in terms of the biogenic, detrital, grain-size proxy,
redox proxy and facies between Unit-1 and Unit-2 are listed. The signature of the
boundary marker is also specified. ............................................................................... 83
Table 3.3: The characteristic differences between sub-unit 1a and 1b in terms of the
biogenic, detrital, grainsize proxy, redox proxy and facies are listed. The
signature of the unit boundary marker is also specified. .............................................. 83
Table 3.4: Thickness of different chemostratigraphic sub-units and units of UBS in all the
eighteen wells used in this study. Well Abv- Abbreviation for well names. ............... 88
Table 4.1: The characteristic differences between sub-units 1a, 1b, 2a and 2b in terms of the
biogenic, detrital, grainsize proxy, redox proxy and facies. ...................................... 128
Table 4.2: Organic geochemistry results for different chemostratigraphic units of Well RT
as derived from the Source Rock Analysis (SRA). Legend: μ: mean; R: Range;
σ: Standard deviation; HI: Hydrogen Index; OI: Oxygen Index; TOCSRA: SRA
measured TOC; TOCorg: Original TOC calculated TOCSRA.................................... 130
Table 4.3: Findings for the paleoredox conditions based on inorganic geochemical methods
and petrographic observations. .................................................................................. 144
Table B.1: Table listing the file description for data used in Chapter-2 of the thesis. ................. 170
Table B.2: Table listing the file description for the data used in Chapter-3 of the thesis. ........... 170
Table B.3: Table listing the file description for the data used in Chapter-4 of the thesis. ........... 171
Table B.4: Table listing the file description of additional data of the Upper and Lower
Bakken Shale collected during the PhD curriculum. ................................................. 171
Table B.5: Table listing the file description for selected published papers and conference
presentations during the PhD curriculum. .................................................................. 171
Table C.1: Table for coefficient of regression (R2) and slope (m) for the best-fit linear
regression lines for the Niton-ICP/Element Analyzer cross plots for eighteen
selected elements showing the relationships between Niton measured elemental

xxi
 concentrations and laboratory-grade measurements made by ICP-MS/ICP-OES
and element analyzer. The XRF-measured concentration of elements which has
been used quantitatively and semi-quantitatively in Chapter 3 are highlighted in
red. ............................................................................................................................. 177
Table C.2: The results for element concentration measurements using Niton handheld XRF
for the repeatability test is listed. The table lists the results for concentration of
trace elements and major element, which were used in this study, for all five runs
for each of the eleven samples (represented by their depth). ..................................... 178

xxii
 4. ACKNOWLEDGEMENTS

This PhD would not have materialized without the sincere support, which I have received in these
5 years, from many individuals. I consider myself privileged to have such a prolific committee, and I take
this opportunity to thank them for sharing their expertise and investing their time in my education.

• Steve Sonnenberg: I will be short of words in thanking you for all your support. Sincere thanks for
entrusting me with all the freedom I enjoyed in driving this research at my own pace and way.
Thanks for always believing in me and providing me with this opportunity. Your dedication and
enthusiasm were a constant source of inspiration, and something, which will keep me motivated
for my entire life.
• Mark Longman: You were always and will be a great source of knowledge and inspiration for me.
Thanks for all your ‘thought-provoking’ questions and insights. Without them, my PhD research
would never have taken this final shape.
• Piret Plink-Björklund: Your Applied Stratigraphy class made me more curious about the
significance of sedimentary processes, which helped me immensely in my PhD research. Thanks
for enriching me with all your inputs on mudrock depositional processes and for always provoking
to think about an alternative interpretation.
• Manika Prasad: You were always inspirational in all our interactions, which were helpful in setting
some of my research goals for this PhD. Thank you for supporting me on giving me direction for
the research on the porosity characterization of the Middle Bakken Member. I would also like to
thank you for giving me the opportunity to present my work at the OCLASSH consortium sponsors
meetings.
• John Curtis: Thanks for agreeing to be on my committee at a very short notice. Thanks for all your
kind suggestions and help, throughout this PhD curriculum.

I sincerely thank all the members and sponsors of the Bakken Consortium for funding this research
and my PhD curriculum at CSM. I would also thank the Bakken Consortium for providing access to data
and samples, which were used in this research. I would also like to extend my gratitude to North Dakota
Geological Survey (NDGS) Grand Forks and US Geological Survey (USGS) core research center Denver
for giving me access to their core libraries and allowing sample collection and data acquisition. I take the
opportunity to remember Late Julie LeFever to express my sincere gratitude for sharing her immense
knowledge about the Bakken Formation. I would like to thank Rocky Mountain Association of Geologists
and Rocky Mountain-SEPM for the scholarships during my PhD. I would also like to thank John Humphrey
for making me interested in carbonates specially in dolomites.

xxiii
 I thank Kathy Emme, Cheryl Medford, and Dorie Chen for providing all the administrative support
during my tenure as a PhD student, including scheduling and planning to meet the thesis timeline. My
colleagues in Bakken Consortium and Geology Department were a great support system with their
cooperation and feedback, thanks to, especially Jingqi Xu, Cosima Theloy, Hui Jin, Mohammad Al
Duhailan, Kazumi Nakamura, Claudia Gutierrez, Gary Listiono.

I also thank my alma-mater Presidency University Kolkata, and Indian Institute of Technology
Roorkee for laying the foundation of my knowledge of Geology. My husband Sanyog Kumar has been
more than just a constant support. Without his motivation and valuable inputs, this PhD would not have
been materialized. I would also thank my father-, mother- and sister-in-law for their immense support.
Finally, I thank my parents, without whom I would not have achieved anything. I would also thank my
daughter Disha for her smiles, which kept me going while I was writing this thesis.

xxiv
To Disha and Sanyog

xxv
 1. CHAPTER 1AA

INTRODUCTION

Bakken Shale Play currently produces with an average rate of 1 million BOPD and contributes
about 23% of the total production from U.S. shale plays. The production boom from Bakken and other
American shale plays is one of the most dramatic growth stories the world of oil has ever seen. The 2017
edition of OPEC’s flagship publication World Oil Outlook (WOO) suggests that non-OPEC supply growth
is dominated by the continuously increasing production from the U.S. shale plays. WOO (2017) further
suggests that shale output is projected to rise from 5.1 million BOPD in 2017 to 7.5 million BOPD by 2021
and to 8.7 million BOPD by 2025. The growth is illustrated in Figure 1.1a, which shows historical U.S.
daily oil production from 1920 to 2017 (data source: EIA); U.S. oil production peaked in October 1970
with an average daily rate of 10.1 MM BOPD, and then declined till 2007. After that year, it witnessed a
steep rise, and daily production is currently near its historical peak in 1970. This remarkable rise is
overwhelmingly driven by production from shale plays, as demonstrated by Figure 1.1b, which provides a
breakdown of daily rate in terms of contribution from the major U.S. shale plays (EIA, 2017). In 2007 these
so-called unconventional plays were contributing less than 10% of U.S. daily rate, whereas in 2017 their
contribution was above 50%. However, despite the success story of U.S. shale play production, WOO
(2017) also raised concerns regarding the possibility of a steep decline in their production after 2025. The
report considered the excessively front-loaded nature of the shale projects, which are primarily driven by
aggressive drilling programs, as the primary reason for this imminent decline. However, given the enormous
in-place volume estimates for the shale resources, such premature decline would essentially leave a large
volume of untapped reserves, which will be unrecoverable with current technology. Therefore, a holistic
improvement in state-of-the-art technology for shale exploration and production is required to enhance
hydrocarbon recovery efficiency. Bridging various knowledge gap in current geoscientific understanding
can immensely improve the efficiency of exploration and production activities in shale plays. This thesis
aims at developing a better understanding of the Bakken Shale Play, by contributing the following: 1)
investigating the significance of the dolomitization process in porosity evolution of the Middle Bakken
Member; 2) defining the lithofacies and chemostratigraphic sub-units of the Upper Bakken Shale; and 3)
understanding the factors controlling the organic richness of the Upper Bakken Shale.

1
Figure 1.1: (a) Chart showing the daily oil production rate of the US from 1920 to 2017 and the historical
peak, which was attained in October 1970. The contribution of shale boom in the daily US oil production
(purple color) post 2000 is also shown. (b) Chart showing a breakdown of the daily production of the
different US shale plays and their relative contribution in the total US production. The chart shows, as of
2017, the Bakken contributes about 23% of the daily shale play production from the US. (Data Source for
charts EIA, 2017)

1.1 Motivation

The Bakken petroleum system consists of two major reservoir intervals: the Middle Bakken
Member (MBM) of the Bakken Formation and the Three Forks Formation. The cumulative estimated
ultimate recoverable (EUR) volume of these two intervals is 5 billion bbl (EIA U.S. Reserves Report, 2016).
The history of oil production from the Bakken Formation dates back to 1953, which marked the discovery
of the Antelope Field. However, the Bakken Formation came into the limelight in 2001, as the hydrocarbon
production potential of the MBM was significantly improved by adopting a strategy of multi-stage
hydraulic fracturing (HF) completion in horizontal wells. The first major success story of this development
was the discovery of the Elm Coulee Field in Montana in 2001. Subsequently, in this field alone, more than
800 such horizontal wells were drilled and completed, the in-place volume is estimated to be 2 billion Bbl,
and the EUR is around 300 MBbl (Theloy, 2014). The Elm Coulee Field has been extensively studied to
understand facies characterization and diagenesis (Sonnenberg and Pramudito, 2009; Alexandre, 2011),
structural interpretations from seismic data (Eidsnes, 2014; Rolfs, 2015), static and dynamic reservoir
modelling (Almanza, 2011) and feasibility studies for Enhanced Oil Recovery (Todd and Evans, 2016).
However, for the Elm Coulee Field, no detailed porosity characterization of the MBM has been performed.
Therefore, for the MBM, there is a lack of understanding of the pore-types, pore-architecture, pore-size, the

2
effect of clay content, and the role of dolomitization, which was the most crucial diagenetic event in terms
of the evolution porosity in MBM.

The naturally fractured Upper Bakken Shale (UBS) was a producing interval from 1976 to 2000
(Sonnenberg, 2014) in the Bakken Play. However, the scale of these production activities was not as
enormous as that of dolomitic siltstone the MBM. This historical production and the recent success in
exploration efforts in UBS in the southwest flank area of the Elm Coulee Field indicate the resource
potential of the UBS (Sonnenberg, 2014). This emphasizes the necessity of developing a better
understanding of the UBS for identification of target intervals and sweet-spots, which will aid in future
exploration and production efforts. The UBS has been extensively studied to understand its depositional
environment, source rock potential, and organic and inorganic geochemistry (e.g. Meissner, 1978; Price et
al., 1984; Webster, 1984; Hayes, 1985; Smith and Bustin, 1996, 1998; Egenhoff and Fishman, 2013; Jin,
2014; Kocman, 2014; Longman et al., 2014). However, there still exist gaps in the understanding of the
UBS with respect to its major lithofacies, and their lateral and vertical distribution, and basin-wide
stratigraphic framework. The organic-rich UBS of the Bakken Formation is a world-class source rock with
an average total organic content (TOC) of 11 wt.% (LeFever et al., 1991; Sonnenberg and Pramudito, 2009;
Jin, 2014). However, there is limited understanding of the factors which have controlled this organic
richness.

1.2 Research Objectives and Methods

The above-mentioned gaps in the understanding of the Bakken Formation have been the motivation
for my research. The following have been my primary research objectives:

1) For the MBM in the Elm Coulee Field, to develop an understanding of the pore-types, pore-
architecture, pore-size, the effect of clay content, and the role of dolomitization in porosity
evolution. To fulfill these research objectives, firstly, the cores were described; this was followed
by thin section petrographic studies, stable isotope analysis, scanning electron microscope (SEM)
imagery, and mercury intrusion pore-size distribution study. The dataset from these analyses and
experiments were integrated with routine core-analysis results and X-Ray diffraction (XRD)
mineralogy data;
2) To identify the lithofacies of the UBS and their lateral and vertical distribution, and to define
laterally correlatable chemostratigraphic units of the UBS, their depositional conditions, and their
thickness variation in the basin. To fulfill these research objectives, firstly, lithofacies were
identified by integrating core description results with thin-section petrographic observations. This
was followed by XRF data acquisition on core slabs for chemostratigraphic analyses. The results
of these two major analyses were integrated to understand the depositional conditions of the UBS;

3
 3) To develop an understanding of the interplay of paleoproductivity, redox condition and dilution of
Organic Matter in controlling the organic richness of the UBS and then to develop a conceptual
model for its deposition. This research objective was fulfilled by integrating the findings of the
ICPMS elemental concentration data, Source Rock Analysis (SRA)-derived Total Organic Content
(TOC), and stable isotopes of carbon and nitrogen.

1.3 Organization of the Thesis

Chapters 2, 3 and 4 document my research as three journal-style papers, and each has an
introduction, sections on geological settings, methodology, results and interpretation, discussion and
conclusion. Chapter 5 summarizes the findings and applications of my research and lists recommendations
for future work. Chapter-2 addresses my first research objective and the title of this paper is “Dolomitization
and its effect on the porosity evolution in the Middle Bakken, Elm Coulee Field.” Chapter-3 focuses on my
second research objective and the title of this paper is “Facies characterization and chemostratigraphy of
Upper Bakken Shale, Williston Basin.” Chapter 4 addresses my third research objective, and the title of this
paper is “Factors controlling organic richness of Upper Bakken Shale, Williston Basin.” General
conclusions and recommendations for future work are provided in Chapter-5.

1.4 Conference Proceedings

I have presented my research findings at various conferences. In addition to the research findings,
I present in this thesis, my presentations have included other contributions on the chemostratigraphy of the
Lower Bakken Shale and the depositional model of the D-Facies of the MBM. Presentations given during
various conference proceedings are as follows:

1) Nandy, D., Sonnenberg, S.A, and Humphrey, J. D. 2014. "Application of inorganic geochemical
studies for characterization of Bakken Shales, Williston Basin, North Dakota and Montana."
Unconventional Resources Technology Conference URTeC: 1922974
2) Nandy, D., Sonnenberg, S. A., and Humphrey, J. D. 2015. Factors Controlling Organic Richness
in Upper and Lower Bakken Shale, Williston Basin: An Application of Inorganic
Geochemistry. Search and Discovery Article #10775.
3) Nandy, D., Listiono, G.A, Sonnenberg, S. A., and Humphrey, J. D. 2015. Mixed-Siliciclastic-
Carbonate System of Facies “D” in the Middle Bakken Formation, Williston Basin. Search and
Discovery Article # 80487.
4) Nandy, D., Kumar, Humphrey, J. D, Longman, M., and Sonnenberg, S. A. 2016. Dolomitization
and its Impact on Porosity Evolution in the Middle Bakken, Elm Coulee Field, Williston Basin.
AAPG 2016 Search and Discovery article

4
 5) Nandy, D., and Sonnenberg, S.A 2017. Petrographic and Geochemical Characterization of the
Upper Bakken Shale, Williston Basin. AAPG 2017 Search and Discovery Article.

1.5 References

Alexandre, C. S., 2011, Reservoir characterization and petrology of the Bakken Formation, Elm Coulee
Field, Richland County, MT: MS thesis, Colorado School of Mines, 175 p.
Almanza, A., 2011, Integrated 3D Geological Model of the Mississippian Devonian Bakken Formation,
Elm Coulee, Williston Basin: Richland County, Montana: MS thesis, Colorado School of Mines, 124
p.
Egenhoff, S. O., and N. S. Fishman, 2013, Traces in the dark-sedimentary processes and facies gradients in
the upper shale member of the Upper Devonian-Lower Mississippian Bakken Formation, Williston
Basin, North Dakota, U.S.A.: Journal of Sedimentary Research, v. 83, no. 9, p. 803–824.
EIA, 2017, Crude Oil Production(Online), Available at: https://www.eia.gov/dnav/pet/pet_crd_crpdn_adc
_mbbl_m.htm
Eidsnes, H. V. H., 2014, Structural and stratigraphic factors influencing hydrocarbon accumulations in the
Bakken Petroleum System in the Elm Coulee field, Williston Basin, Montana: Colorado School of
Mines, 137 p.
Hayes, M. D., 1985, Conodonts of the Bakken Formation (Devonian and Mississippian), Williston Basin,
North Dakota: The Mountain Geologist, v. 22, no. 2, p. 64–77.
Jin, H., 2014, Source rock potential of the Bakken shales in the Williston Basin, North Dakota and Montana:
PhD Thesis, Colorado School of Mines, 220 p.
Kocman, K. B., 2014, Interpreting Depositional and Diagenetic Trends in the Bakken Formation Based on
Handheld X-Ray Fluorescence Analysis, Mclean, Dunn, and Mountrail Countries, North Dakota: MS
thesis, Colorado School of Mines, 156 p.
LeFever, J. A., C. D. Martiniuk, E. F. R. Dancsok, and P. A. Mahnic, 1991, Petroleum potential of the
middle member, Bakken Formation, Williston Basin: in J. E. Christopher and F. Haidl, eds.,
Proceedings of the Sixth International Williston Basin Symposium: Saskatchewan Geological Society,
Special publication 11, p. 74-94.
Longman, M., K. Kocman, and L. Wray, 2014, Petrography of the Bakken black “shales” in the eastern
Williston Basin of North Dakota: AAPG Datapages/Search and Discovery Article #90193, Rocky
Mountain Section AAPG Annual Meeting, Denver, Colorado, July 20–22.
Meissner, F. F., 1978, Petroleum geology of the Bakken Formation, Williston Basin, North Dakota and
Montana, in D. Rehrig, eds., The economic geology of the Williston Basin: Proceedings of the
Montana Geological Society, 24th Annual Conference, p. 207-227.
WOO, Organization of the Petroleum Exporting Countries, 2017, OPEC World Oil Outlook, Available
from: http://www.opec.org.
Price, L. C., T. Ging, T. Daws, A. Love, M. Pawlewicz, and D. Anders, 1984, Organic Metamorphism in
the Mississippian-Devonian Bakken Shale, North Dakota Portion of the Williston Basin: in J.
Woodward, F. F. Meissner and J. L. Clayton, eds., Hydrocarbon Source Rocks of the Greater Rocky
Mountain Region, Rocky Mountain Association of Geologists, p. 83–133.
Rolfs, S. A., 2015, Integrated geomechanical, geophysical, and geochemical analysis of the Bakken
Formation, Elm Coulee field, Williston Basin, Montana: MS thesis, Colorado School of Mines, 122
p.

5
Smith, M. G., and R. M. Bustin, 1996, Lithofacies and paleoenvironments of the Late Devonian and Early
Mississippian Bakken Formation, Williston Basin: Bulletin of Canadian Petroleum Geology, v. 44,
no. 3, p. 495–507.
Smith, M. G., and R. M. Bustin, 1998, Production and preservation of organic matter during deposition of
the Bakken Formation (Late Devonian and Early Mississippian), Williston Basin: Palaeogeography,
Palaeoclimatology, Palaeoecology, v. 142, no. 3–4, p. 185–200.
Sonnenberg, S., 2014, The Upper Bakken Shale Resource Play, Williston Basin: Society of Exploration
Geophysicists, American Association of Petroleum Geologists, Society of Petroleum Engineers,
Proceedings of the 2nd Unconventional Resources Technology Conference, Denver, Colorado, 25–27
August, p. 189-200.
Sonnenberg, S. A., and A. Pramudito, 2009, Petroleum geology of the giant Elm Coulee field, Williston
Basin: AAPG Bulletin, v. 93, no. 9, p. 1127–1153.
Theloy, C., 2014, Integration of Geological and Technological Factors Influencing Production in the
Bakken Play, Williston Basin: PhD Thesis, Colorado School of Mines, 241 p.
Todd, H. B., and J. G. Evans, 2016, Improved Oil Recovery IOR Pilot Projects in the Bakken Formation,
in SPE Low Perm Symposium: Society of Petroleum Engineers.
Webster, R. L., 1984, Petroleum source rocks and stratigraphy of the Bakken Formation in North Dakota:
in Woodward J., F. F. Meissner, J. L. Clayton, eds., Hydrocarbon source rocks of the greater Rocky
Mountain region, Rocky Mountain Association of Geologists, p. 57 -82.

6
 2. CHAPTER 2

DOLOMITIZATION AND ITS EFFECT ON POROSITY EVOLUTION IN MIDDLE

BAKKEN, ELM COULEE FIELD

The current state of the art on the characterization of sub-micron sized pores in tight carbonate
rocks are mostly limited to limestones and chalks. But sub-micron sized pores are also common and are of
significance in dolostones since they affect the hydrocarbon saturation and production performance. In this
paper, I present a case study on the microporosity characterization of the Middle Bakken silty dolomites of
the Elm Coulee field, Williston Basin. This paper also develops an understanding of the various
dolomitization stages, which affects the evolution of porosity in the Middle Bakken silty dolostones. The
shallow marine mixed-siliciclastic-carbonate deposits of the Middle Bakken Member (MBM) in Elm
Coulee Field are dolomitized to varying degrees, and the dolomite content ranges from 5 to 65 wt.%.
Dolomitization process in MBM of Elm Coulee Field was defined by a collective analysis of petrographic
observation, X-Ray Diffraction (XRD) mineralogy, and stable isotope of C and O results of the core samples
from eight wells of the field. Moreover, routine core analysis, Mercury Intrusion Porosimetry (MIP), and
Field Emission Scanning Electron Microscope (FESEM) studies were performed on these core samples to
determine the total porosity, pore-size distribution, and pore architecture.

Dolomite crystals with planar subhedral to euhedral fabric, stable isotope results, the presence of
late anhydrite cement and overdolomitization indicate that the Middle Bakken dolostones in Elm Coulee
Field are the product of seepage-reflux dolomitization in mesohaline conditions. Relationships between
dolomite content, total porosity, and pore size suggest that the porosity evolution in the MBM is closely
linked with the dolomitization process and is mainly dependent on the degree of dolomitization. A
progressive increase in the dolomite content (up to 60 wt.%) results in an overall increase in the
development of intercrystalline pores, total porosity, pore size and dolomite crystal size. However, with
further increase in dolomite content beyond 60 wt.%, (overdolomitization), total porosity and pore size
decrease as the size of the dolomite crystals increases. In the MBM, the dolomitization process involves
volume-for-volume replacement of precursor calcite by dolomite in an open system. Moreover, the degree
of dolomitization is dependent on the external supply of Mg2+and (CO3)2- ions. In this process, the total
porosity remains same, while the pores size increases. However, as limestone is more susceptible to
compaction than dolomite, so higher dolomite content in the MBM results in a comparatively lower porosity
loss due to compaction. This explains the increase in porosity in the MBM with an increase in dolomite
content. However, in the overdolomitization stage, dolomite crystals form interlocking fabrics, which in
turn decreases the total porosity and pore size due to the formation of dolomite cement overgrowths.

7
2.1 Introduction

About a half of the world’s carbonate type petroleum reservoirs are dolostones (Al-Awadi et al.,
2009); however, there are no modern-day analogs to the ancient massive dolostone systems. Therefore, the
origin of dolostones and the process of dolomitization have been extensively studied in the past.
Dolomitization commonly results in reservoirs with unique geometries and porosity distribution, which
have a direct impact on exploration and production strategies. The role of dolomitization in porosity
evolution was discussed by Powers (1962); Davies (1979); Allan and Wiggins (1993); Sun (1995); Warren
(2000); Lucia (2004); Ehrenberg et al. (2006); Choquette and Hiatt (2008); Maliva et al. (2011); and Wang
et al. (2015). After even five decades of research, the topic of the origin of porosity in dolostones is still
controversial (Purser et al., 1994; Lucia, 2004; Ehrenberg et al., 2012). It has been observed that deeply
buried, particularly Paleozoic dolostones, have better porosity than the associated limestone (Landes and
Arbor, 1946; Murray, 1960; Longman et al., 1983; Schmoker et al., 1985; Purser et al., 1994; Sun, 1995;
Saller and Henderson, 1998). However, because of this observation, it has been postulated that the
dolomitization process results in an increase in porosity. This hypothesis gained further support in its favor
due to the mole-for-mole replacement equation of dolomitization, which is 2CaCO3 + Mg2+= CaMg(CO3)2
+Ca2+. De Beaumont (1837) suggested that the mole-for-mole conversion of calcite to dolomite results in a
12.96% decrease in grain volume because the molar volume of dolomite is less than the molar volume of
calcite. This contraction of grain volume is believed to cause a 12.96% increase in porosity of the
dolomitized rock. Despite a multitude of case studies showing relatively higher porosity in dolostones, there
are examples where a contrary relationship has been reported. For example, Lucia and Major (1994)
observed that for the Plio-Pleistocene carbonates in Bonaire, Netherlands Antilles, the precursor limestones
are more porous than the corresponding dolostones. Similarly, in their description of the upper Paleozoic
buildup reservoir from the Horseshoe Atoll complex, Texas, Saller et al. (2004) noted that dolostones have
lower porosity than the associated limestones. Furthermore, the Upper Jurassic Arab reservoirs of Saudi
Arabia and Qatar show an overall decrease in porosity with dolomitization (Powers, 1962; Cantrell et al.,
2004). Maliva et al., (2011) reported that the dolostones and associated limestones have nearly the same
total porosity, thereby showing no change in porosity due to dolomitization. Therefore, the literature
suggests that the porosity in dolostones can be enhanced or reduced, or it can remain same due to the process
of replacement of calcite in limestones by dolomite. This ambiguity was explained by Morrow (1982) on
the basis of the following volume-for-volume reaction: (2-x)CaCO3 + x(CO3)2- + Mg2+= CaMg(CO3)2
+Ca2+, where x=moles of external carbonate (CO3)2- ions added from the dolomitizing fluid. Therefore,
depending on the amount of external (CO3)2- ions supplied to the system by the dolomitizing fluid, porosity
of the resulting dolostone may increase, remain constant or decrease. Lucia (2004) suggested that

8
dolomitization does not create additional porosity; however, the external (CO3)2- ions in the solution results
in an occlusion of porosity.

This study focuses on the hydrocarbon bearing interval of the Middle Bakken Member (MBM)
silty dolostones, in the Bakken Formation of the Elm Coulee Field, Montana. The porosity of the MBM
ranges from 2% to 10% in the Elm Coulee Field, while the permeability varies from 0.02-0.04mD
(Sonnenberg and Pramudito, 2009; O’Brien et al., 2012). Sonnenberg and Pramudito (2009) and Alexandre
(2011) observed that the MBM in those sub-intervals and location, which are dolomite-rich, have better
reservoir quality (porosity and permeability) and better hydrocarbon saturation. However, a detailed
analysis of porosity, which is shown in this study, suggests that this may not always be true in all cases.
Moreover, Alexandre (2011) has observed micropores in petrographic thin sections of the Middle Bakken
from the Elm Coulee Field under an epifluorescence microscope. The term “micropores” from here onward
refers to less than 1µm throat-size pores, according to the convention used by Pittman (1971) and Lucia
(1995). However, by definition from Loucks et al., (2012), these sub-micron sized pores would be classified
as nanopores, which is the convention of International Union of Pure and Applied Chemistry (IUPAC).
However, recent studies focused on the evolution of sub-micron sized pores in carbonates are limited to
limestones and chalks (Deville de Periere et al., 2011; Lucia and Loucks, 2013; Kaczmarek et al., 2015;
Loucks and Ulrich, 2015). This study gives the opportunity to see the effect of dolomitization in the
evolution of micropores in the Middle Bakken. Moreover, this study provides an explanation of these
observations through the effect of dolomitization on different aspects of porosity evolution in the MBM.

Investigation of petrography and corresponding measured porosity and pore size distribution of
carbonates, especially in the transition zone from limestone to dolostone in cores, could link the textural
evolution to the porosity changes during the dolomitization process. In this chapter, I summarize previous
studies on the formation mechanisms of porosity in dolostone and then investigate how the process of
dolomitization affects porosity evolution in the MBM in the Elm Coulee Field by integrating core
descriptions, petrography, core analysis and stable isotope data. The focus of this study is to understand the
changes in texture, pore structure, porosity and pore size that occur through different stages of
dolomitization.

The main objectives of this study with respect to the MBM in the Elm Coulee Field are:

1. Characterize the facies of the MBM from core description

2. Identify diagenetic stages based on petrographic studies
3. Identify dolomitization model based on petrographic and geochemical constraints
4. Characterize porosity both qualitatively and quantitatively
5. Investigate how the process of dolomitization affects porosity evolution in the MBM

9
2.2 Literature review on origin of porosity in dolostones

Dolostones are formed as a result of replacement of calcite in limestones by dolomite. Therefore,

the origin of porosity in dolostones can be divided into three categories: pre-dolomitization,
syndolomitization, and post-dolomitization, based on the relationship between the timing of porosity
formation and the dolomitization process. When porosity in the dolostone is inherited from the precursor
limestone and do not change due to dolomitization, it is termed as the pre-dolomitization origin of porosity.
When porosity, pore space or pore size is altered during dolomitization, it is termed as the syndolomitization
origin of porosity. When total porosity, pore types and pore size is altered after dolomitization and is
controlled by fluids unrelated to the dolomitization process, it is termed as the post-dolomitization origin
of porosity. The syndolomitization origin of porosity has been described in the literature either by the mole-
for-mole replacement reaction or by the volume-for-volume replacement reaction. These two reactions are
described in the section below.

Mole-for-mole replacement
When dolomite replaces calcite, Mg2+ ions substitutes half of the Ca2+ ions from the crystal lattice.
The reaction is expressed by the following equation (Tucker and Wright, 1990):

2CaCO3 + Mg2+ = CaMg(CO3)2 +Ca2+ (2.1)

De Beaumont (1837) proposed that the conversion of limestone to dolostone is based on this classic
mole-for-mole replacement reaction, which causes an overall shrinkage in grain volume as the molar
volume of Mg2+ is less than that of Ca2+. The decrease in grain volume, in turn, increases porosity in the
dolostones. This theory has been extensively used to interpret the formation of porous dolostone, especially
in deeply buried carbonate strata where dolostones are observed to be more porous than the surrounding
limestones. However, the mole-for-mole replacement reaction is countered by many authors (Morrow,
1982; Land, 1983; Lucia, 2004) as it assumes that dolomitization takes place in a closed system, in which
only Mg2+ are added to the system from the dolomitizing fluid. Weyl (1960) countered this by
demonstrating that (CO3)2- ions are present in a minor fraction in the formation waters with respect to Mg2+
and Murray (1960) suggested that (CO3)2- ions involved in the dolomite precipitation are mainly derived
from local carbonate sources. The mole-for-mole mechanism also fails to explain the occurrence of lower
porosity in dolostones compared to the associated limestone. For example, in Plio-Pleistocene carbonates
of Bonaire limestones and dolomitic limestones have similar ranges of 20–30% porosity, whereas the
adjacent dolostones have the lower porosities of about 11% (Lucia and Major, 1994), which cannot be
explained by the mole-for-mole replacement reaction. This also fails to account for the occurrence of
dolomitic pore occluding cement. Such contrary observations suggest that external (CO3)2- ions must be
added to the replacement reaction during the dolomitization process to form the extra dolomite. Moreover,

10
most dolomitization processes are supposed to take place in open to semi-open systems (Land, 1983); the
dolomitizing fluid which brings Mg2+ ions can also bring in (CO3)2- ions and remove the Ca2+ ions.
Moreover, the Ca2+ ions released from the precursor calcite need to be taken out of the reaction site to
maintain the optimum Mg/Ca ratio required for precipitation of dolomite.

Volume-for-Volume Reaction
Many studies in the past have observed that the porosity remains unchanged as limestone is
dolomitized (Maliva et al., 2011). This was explained by Morrow (1982) based on the following volume-
for-volume replacement reaction:

(2-x) CaCO3 + x(CO3)2- + Mg2+= CaMg(CO3)2 +(1-x) Ca2+ (2.2)

Where x is the number of moles of external (CO3)2- ions added from the dolomitizing fluid. It is proposed
that the volume of (CO3)2- ions required to replace the difference in molar volume is 0.1 for aragonite and
0.25 for calcite. Lucia (2004) suggests that in this condition of dolomitization, although the total porosity
in dolostones is inherited from the precursor limestone, dolomitization can modify the pore space and
thereby change the pore size. After all the precursor calcite has been replaced, but the dolomitizing fluid
have optimum Mg2+ and (CO3)2-ions, then dolomite will continue to precipitate. This results in a decrease
in porosity and pore throat size and is termed as “overdolomitization” by Halley and Schmoker (1983).
Maliva et al., (2011) suggested that dolomitization affects porosity differentially on the scale of
individual crystals, beds, and formations. At the crystal scale, dolomitization occurs by volume-for-volume
replacement reaction in which the space occupied by the precursor calcite is replaced by a solid dolomite
rhomb with an associated reduction in porosity. However, as (CO3)2- ions are passively scavenged from the
formation, so there is no change in porosity at formation scale. However, in bed scale, the creation of moldic
pores from allochems depends on the relative rate of calcite dissolution and the rate of dolomite
precipitation.

2.3 Geological Setting

The Elm Coulee Field is located along the southwestern margin of the Williston Basin within
Richland County in northeast Montana (Figure 2.1). Williston Basin is an elliptical intracratonic basin
located along the western edge of the Canadian Shield and covering an area of around 10,000 square miles
(Kent and Christopher, 1994). The basin is likely to have originated as a craton-margin basin and evolved
into an intracratonic basin during the Cordilleran orogen (Gerhard et al., 1990). The present-day basin
configuration is the result of the Laramide orogeny during Late Cretaceous and early Tertiary time (Kent
and Christopher, 1994). A continuous sediment column about 16,000ft thick accumulated in the Williston
Basin from the Cambrian to the Tertiary age (LeFever et al., 1991). Cyclical transgressions and regressions

11
characterize sedimentation in the basin with repeated deposition of carbonates and siliciclastics with many
unconformities. Paleozoic strata consist mainly of cyclic carbonate deposits while Mesozoic and Cenozoic
strata consist mainly of siliciclastics. Initial sedimentation during the Cambrian time started over an
irregular Precambrian surface. The basin began to subside during the Ordovician Period and underwent
episodic subsidence throughout the rest of the Phanerozoic period.

During the Late-Devonian-Early Mississippian age, the Bakken Formation was deposited in the
Williston Basin. During its deposition, the basin was an area of active subsidence. The Three Forks
Formation unconformably underlies the Bakken Formation and is overlain by the Lodgepole Formation
(Figure 2.2). The Bakken Formation has been divided into three members, the organic-rich Upper and
Lower Bakken Shale, which are the source rocks, and the silty carbonates of the MBM, which is the main
reservoir interval. Previous studies suggest that the organic-rich Upper and Lower Bakken shale was
deposited in an offshore marine anoxic environment during periods of sea level rise (Price et al., 1984;
Webster, 1984; Smith and Bustin, 1998). The MBM shows a wide range of facies variability and is
interpreted to have been deposited in a shallow marine environment in an epeiric platform setting following
a rapid sea level drop, resulting in a regressive event (Webster, 1984; LeFever et al., 1991; Smith and
Bustin, 1996; Angulo and Buatois, 2012). In the Elm Coulee Field, the MBM ranges in thickness from 10ft
to 40ft and consists of silty lime wackestones to bioturbated siltstones, parts of which have been dolomitized
to varying degrees. The lithology ranges from pure silty limestone to partially dolomitized silty limestones
and pure silty dolostones (Sonnenberg and Pramudito, 2009). The Lower Bakken Shale approaches its
erosional zero edge in the Elm Coulee Field (Sonnenberg and Pramudito, 2009; Alexandre, 2011; Rolfs,
2015).

2.4 Methodology and Dataset

Samples for this study were obtained from the cores of the MBM interval from the Elm Coulee
Field wells as shown in Figure 2.3. Enerplus, which is the operator of the field, provided these cores along
with petrographic thin sections, routine core analysis (RCA) data and X-ray diffraction (XRD) mineralogy
from eight wells to the Bakken Consortium at Colorado School of Mines. Details of the wells, including
well name, API, Township, Section and Range and available data including cores, thin sections, RCA and
XRD used in this study is given in Table 2.1. Detailed core description was performed on all the eight cores
for lithofacies identification. Sedimentary facies were described based on lithology, sedimentary structures,
trace-fossil, and bioturbation index (BI). BI was estimated based on the scheme proposed by Taylor and
Goldring (1993). Core plugs from the eight Enerplus cores, which were taken at a 1ft interval, were sent to
Core Laboratories Inc. for RCA. The Core Measurement System-300 (CMS) was used to measure porosity
and permeability. Mineralogy for these samples was obtained using XRD at Mineral Lab, Golden,

12
Colorado. The end trims of the core plugs were used for the preparation of the petrographic thin sections.
Thin sections were impregnated with blue epoxy and stained with alizarin red and potassium ferrocyanide
to distinguish calcite, dolomite, and ferroan-dolomite. Half of the thin sections were injected with a pink
epifluorescent dye to highlight pore geometries when exposed to a fluorescent light under a microscope. A
total of two hundred and twenty-four petrographic thin sections have been examined in this study.

Figure 2.1: Location map of the Williston Basin with subsea structure contour on the base of the
Mississippian. Elm Coulee Field is located in northeast Montana, demarcated by the black ellipse . The
limit of the Bakken Formation is shown by a dashed line. P=location of the Parshall field area; A=location
of Antelope field.(modified from Sonnenberg and Pramudito, 2009).

Figure 2.2: Late Devonian-Early Mississippian stratigraphic column of the Williston Basin showing the
Bakken Formation and the underlying Three Forks Formation and overlying Lodgepole Formation.
(modified from Webster (1984) and Sonnenberg and Pramudito (2009)).

13
 From the large inventory of core plug samples, twenty samples representing the range of porosity
and dolomite content were selected for Field-Emission-Scanning Electron Microscopy (FE-SEM) in order
to examine the pore architecture and identify the pore types. These twenty samples were examined under
JEOL JSM-7000F FE-SEM at Colorado School of Mines. Pore throat-size distribution was obtained for
seven representative samples using Mercury Intrusion Porosimetry (MIP). The opposite core plug end-trim
was sent to Micromeretics Analytical Services for MIP analysis. Mercury intrusion and extrusion data were
collected on the AutoPore IV 9500 V1.09 machines on sample masses ranging from 2-5 grams. Pore radius
was calculated from the volume of mercury intrusion. Stable isotope of oxygen (O) and carbon (C) of twelve
representative pure dolostone samples were analyzed from the powdered samples of whole rock, collected
from the end trims of the core plugs by using a drill. It was ascertained through XRD mineralogy and thin
section petrography that the samples selected for isotope analysis were pure dolostones and did not have
calcite as a contaminant. Further details about sample selection for stable isotope analysis is discussed in
Section 2.5.3. Stable isotope analysis was conducted on about 200 µg of powder were collected for each
sample and were analyzed in Thermo Finnigan GasBench II coupled to a Thermo Delta Plus XL at the
Stable Isotope Facility at University of Wyoming, Laramie. Oxygen and carbon isotope of carbonates are
reported in per mil (‰) relative to the PDB (Pee Dee Belemnite) standard.

Table 2.1: Name and location of the wells from the Elm Coulee Field from which cores, thin sections,
routine core analysis (RCA) and X-ray diffraction (XRD) mineralogy are used in this study.
Well Name API T/S Range Sec. Core Thin RCA XRD
Section
RR Lonetree Edna 1-13 LT 250832269500 T23N R56E 1 ✓ ✓ ✓ ✓
Coyote Putnam CP 250832283700 T23N R57E 9 ✓ ✓ ✓ ✓
Bullwinkle Yahoo 4-1- BY 250832273600 T24N R57E 4 ✓ ✓ ✓ ✓
HSU
Stockade Jayla 32-3- SJ 250832273700 T25N R51E 32 ✓ ✓ ✓ ✓
HID3
Peanut Jimmy 22-3- PJ 250832243200 T24N R57E 22 ✓ ✓ ✓ ✓
HID3
Jackson Rowdy 3-8 JR 250832279300 T26N R51E 3 ✓ ✓ ✓ ✓
Brutus East Lewis 3-4-H BE 250832250700 T24N R57E 3 ✓ ✓ ✓ ✓
L
Foghorn Ervin 22-3-H FE 250832273900 T23N R58E 20 ✓ ✓ ✓ ✓

2.5 Results and Interpretation

This section, firstly, details the various lithofacies of MBM of Elm Coulee Field based on the core
description. Thereafter, the major diagenetic events for the MBM is described. This section also provides
the details of the mechanism of dolomitization in the MBM. Finally, the relationship between dolomite
content and total porosity, pore size, and pore architecture is outlined.

14
Figure 2.3: Isopach map of the Middle Bakken Member with the location of the wells studied in this paper.
The extent of the Elm Coulee Field is outlined by the green line. The red line is the boundary of the Richland
County. Location of the wells which were studied for core description and routine core analysis (RCA)
and X-ray diffraction (XRD) derived mineralogy were examined are shown by the yellow circles.
Abbreviation of the well names shown in this map is given in Table 2.1.

Lithofacies Characterization

The Middle Bakken facies identified in this study in the Elm Coulee region, on the basis of
lithology, sedimentary structures, trace fossils, and BI are similar to the facies described by previous authors
(LeFever et al., 1991; Pramudito, 2008; Simenson, 2011; Gent, 2011; Angulo and Buatois, 2012). Nine
sedimentary facies were identified from the cores of the MBM in the Elm Coulee Field. However, these
nine sedimentary facies were grouped into five major facies, which from bottom to top are: A, B, D, E, and

15
F, in order to be consistent with the facies nomenclature scheme established by the Bakken Consortium at
the Colorado School of Mines. Moreover, facies C of the MBM is not developed in the Elm Coulee Field
region. Details of this classification scheme are referred from Gent (2011) and Simenson (2011).Table 2.2
lists a detailed description of the facies identified in the MBM of the Elm Coulee region. Representative
core photographs of each of the nine facies are shown in Figure 2.4. Core description, along with the depth
profile of the mineralogy obtained from XRD analysis and porosity and saturation data obtained from RCA
of the Coyote Putnam well representing the MBM in the Elm Coulee Field, is shown in Figure 2.5. Core
description panels of other seven wells are included in Appendix-A. Within an individual location in the
field, such as the cored interval of the well Coyote Putnam (Figure 2.5), the dolomite content varies
considerably along depth and between different facies of the MBM. There is an overall increase in the
dolomite content and a decrease in calcite content from the bottom to the top of the Middle Bakken facies
B. Moreover, there is a considerable variation in the abundance and distribution of dolomite and calcite
content in the MBM, both spatially and temporally.

Diagenesis

A detailed petrographic examination was conducted to understand the diagenesis in Middle Bakken
in the Elm Coulee Field. The diagenetic processes identified in the Middle Bakken in the Elm Coulee Field
are mechanical and chemical compaction, early calcite cementation, early dolomitization, formation of
pyrite, authigenic illite clay, quartz overgrowth and replacement, anhydrite cementation, minor ankerite
(ferroan) dolomite overgrowth, fracturing and hydrocarbon production. Figure 2.6 shows the paragenetic
sequence of the Middle Bakken in the Elm Coulee Field. Alexandre (2011) studied diagenesis and
developed a paragenetic sequence for the MBM in the Elm Coulee Field. This study supports most of the
diagenetic stages recognized by Alexandre (2011). However, this study contradicts the “dedolomitization”
stage identified by Alexandre (2011). Alternatively, patchy calcite cementation is identified in the Middle
Bakken in the Elm Coulee Field based on core description and petrographic analysis (Figure 2.7a & b).
Prior to calcite cementation, bioclasts were subjected to micritization as evident from the presence of
micritization envelope that developed on bioclasts (Figure 2.7c). Micritization indicates diagenesis in
marine condition. Euhedral dolomite rhombs formed within calcite cement indicates that calcite
cementation was followed by dolomitization. Similar early patchy calcite cementation that predates
dolomitization has also been identified in the Middle Bakken in other parts of the Williston Basin (Brennan,
2016). Brennan (2016), based on stable isotope data from the Middle Bakken in North Dakota,
demonstrated that the original source of calcium carbonate in these patchy calcite cement was most likely
marine carbonate debris mixed internally throughout the member, which dissolved and reprecipitated as

16
calcite cement. The author also concluded that the dissolution of the internal carbonate has most likely
taken place in the meteoric realm within the shallow subsurface.

Table 2.2: Sedimentary and ichnologic characteristics of the sedimentary facies identified in the Middle
Bakken in the Elm Coulee Field.
Facies Lithology Sedimentary structures Bioturbation Ichnofossils
Index
F Light gray, whole fossils bearing Massive, burrow-mottled, 0-1 burrow
brachiopod shells and crinoid storm beds present mottling
dolomitic-to lime wackestone
E2 Dark-gray, very thinly Parallel lamination; locally 0-1 Planolites
interlaminated, muddy siltstone, current ripple and mudstone montanus
may be dolomitic or calcareous drapes
E1 Light- to dark-gray and buff Wavy lamination; mudstone 0
colored, commonly pyritic in drapes; microfaults and soft
places, calcareous to dolomitic sediment deformation occur
siltstone occasionally
D2 Light gray, oolitic grainstone- Massive to cross stratified, 0
sandstone occasionally wavy lamination
present
D1 Light gray, fine-grained silty Flaser-bedded with current 0
dolostone, mud drapes commonly ripples; climbing ripples
present and mudstone drapes are also
common
B3 Interbedded massive, light-gray to Massive with common intervals 3-4 Planolites
buff colored fine-grained silty of wavy or parallel lamination, montanus
dolostone with muddy partings continuous shale laminae occur
B2 Interbedded light-gray to buff Bed boundaries are diffuse, 4 Nereites
colored, massive, fine-grained locally continuous shale missouriensis
silty dolostone, patchy calcite laminae occur and minor
cementation present. Phycosiphon
incertum and
Planolites
montanus
B1 Light-gray or greenish-gray Massive. Very rarely 5 Phycosiphon
burrow-mottled calcareous to microhummocky and very thin incertum and
partially dolomitized siltstone parallel lamination occurs in Nereites
, pyritic, brachiopod shell remains the siltier intervals missouriensis
and discontinuous thin laminae of
shale
A Greenish-gray burrow-mottled, Massive with burrow-mottled Phycosiphon
very fine grained silty limestone, texture incertum
with fragments of brachiopod
shells and crinoids may be
dolomitic

This was followed by dolomitization of the precursor micritic lime mud. The Middle Bakken
dolostones predominantly consist of planar subhedral to euhedral dolomite rhombs with unreplaced (rarely
partially) allochems (Figure 2.8a & b). Zonation within the dolomite rhombs (Figure 2.8b) is also observed,
which indicates slight changes in the chemistry of the dolomitizing fluid. Alexandre (2011) mentioned the

17
presence of high-temperature saddle dolomite in minor quantities in the Middle Bakken in the Elm Coulee
Field. However, saddle dolomite was not observed in any of the thin sections examined in this study. Late
stage ankerite (ferroan dolomite) was observed only in very few samples, as thin rims of dolomite rhombs.
Further details about dolomitization are discussed in the Sections 2.5.3 and 2.6.1. Pyrite occurs in many
different forms throughout the diagenetic sequence. It is found both in the center and around the edges of
the dolomite rhombs (Figure 2.9a). It replaces fossils (Figure 2.9b-d), occasionally fills burrows and
fractures, and also forms small framboids. Therefore, pyrite apparently began to form very early in the
sequence and continued until the hydrocarbon generation process effectively stopped diagenesis.

Figure 2.4: The nine different lithofacies identified in the Middle Bakken Member (MBM) of the Bakken
Formation in the Elm Coulee Field are shown in the core photographs. Detail description of each facies is
listed in Table 2.2.

18
Figure 2.5: Core description of well Coyote Putnam showing lithofacies variation in the Middle Bakken
along with the corresponding depth profiles of porosity and saturation obtained from routine core analysis
and dolomite, calcite and clay content obtained from XRD analysis. MB-Middle Bakken, UBS-Upper
Bakken Shale. Details about the degree of dolomitization is discussed later in Section 2.5.5.

19
 After the dolomitization process ceased, authigenic quartz (Figure 2.10a), illite (Figure 2.10b) and
secondary anhydrite (Figure 2.11 b-f) precipitated as cement and overgrowth. Silica and anhydrite
replacement of burrows was also observed (Figure 2.11a). Formation of authigenic illite in the Middle
Bakken was not recognized by Alexandre (2011). Pitman et al. (2001) suggested that the pore water became
supersaturated with potassium due to the dissolution of K-feldspar framework grains, which precipitated as
authigenic illite in the Middle Bakken. Finally, diagenesis in the Middle Bakken ended after hydrocarbon
expulsion and formation of fractures. The continuation of hydrocarbon formation increased pressures within
the shales, eventually causing them to fracture and expel hydrocarbons into the Middle Bakken reservoir
(Pitman et al., 2001) .

A detailed petrographic examination was conducted to understand the diagenesis in Middle Bakken
in the Elm Coulee Field. The diagenetic processes identified in the Middle Bakken in the Elm Coulee Field
are mechanical and chemical compaction, early calcite cementation, early dolomitization, formation of
pyrite, authigenic illite clay, quartz overgrowth and replacement, anhydrite cementation, minor ankerite
(ferroan) dolomite overgrowth, fracturing and hydrocarbon production. Figure 2.6 shows the paragenetic
sequence of the Middle Bakken in the Elm Coulee Field. Alexandre (2011) studied diagenesis and
developed a paragenetic sequence for the MBM in the Elm Coulee Field. This study supports most of the
diagenetic stages recognized by Alexandre (2011). However, this study contradicts the “dedolomitization”
stage identified by Alexandre (2011). Alternatively, patchy calcite cementation is identified in the Middle
Bakken in the Elm Coulee Field based on core description and petrographic analysis (Figure 2.7a & b).
Prior to calcite cementation, bioclasts were subjected to micritization as evident from the presence of
micritization envelope that developed on bioclasts (Figure 2.7c). Micritization indicates diagenesis in
marine condition. Euhedral dolomite rhombs formed within calcite cement indicates that calcite
cementation was followed by dolomitization. Similar early patchy calcite cementation that predates
dolomitization has also been identified in the Middle Bakken in other parts of the Williston Basin (Brennan,
2016). Brennan (2016), based on stable isotope data from the Middle Bakken in North Dakota,
demonstrated that the original source of calcium carbonate in this patchy calcite cement was most likely
marine carbonate debris mixed internally throughout the member, which dissolved and reprecipitated as
calcite cement. The author also concluded that the dissolution of the internal carbonate has most likely
taken place in the meteoric realm within the shallow subsurface.

This was followed by dolomitization of the precursor micritic lime mud. The Middle Bakken
dolostones predominantly consist of planar subhedral to euhedral dolomite rhombs with unreplaced (rarely
partially) allochems (Figure 2.8a & b). Zonation within the dolomite rhombs (Figure 2.8b) is also observed,
which indicates slight changes in the chemistry of the dolomitizing fluid. Alexandre (2011) mentioned the

20
presence of high-temperature saddle dolomite in minor quantities in the Middle Bakken in the Elm Coulee
Field. However, saddle dolomite was not observed in any of the thin sections examined in this study. Late
stage ankerite (ferroan dolomite) was observed only in very few samples, as thin rims of dolomite rhombs.
Further details about dolomitization are discussed in the Sections 2.5.3 and 2.6.1. Pyrite occurs in many
different forms throughout the diagenetic sequence. It is found both in the center and around the edges of
the dolomite rhombs (Figure 2.9a). It replaces fossils (Figure 2.9b-d), occasionally fills burrows and
fractures, and also forms small framboids. Therefore, pyrite apparently began to form very early in the
sequence and continued until the hydrocarbon generation process effectively stopped diagenesis.

After the dolomitization process ceased, authigenic quartz (Figure 2.10a), illite (Figure 2.10b) and
secondary anhydrite (Figure 2.11 b-f) precipitated as cement and overgrowth. Silica and anhydrite
replacement of burrows was also observed (Figure 2.11a). Formation of authigenic illite in the Middle
Bakken was not recognized by Alexandre (2011). Pitman et al. (2001) suggested that the pore water became
supersaturated with potassium due to the dissolution of K-feldspar framework grains, which precipitated as
authigenic illite in the Middle Bakken. Finally, diagenesis in the Middle Bakken ended after hydrocarbon
expulsion and formation of fractures. The continuation of hydrocarbon formation increased pressures within
the shales, eventually causing them to fracture and expel hydrocarbons into the Middle Bakken reservoir
(Pitman et al., 2001) .

Figure 2.6: The paragenetic chart is showing the different diagenetic events that the Middle Bakken
Member (MBM) has undergone in the Elm Coulee Field.

21
Figure 2.7: (a) The core photograph is showing patchy calcite cementation (white dashed line) in the Middle
Bakken facies B. CP, 10365.2ft. (b) The photomicrograph is showing the patchy calcite (red) cement in the
Middle Bakken facies B. Plane light, BEL,10411.3ft. (c) The photomicrograph is showing micritization
envelope which developed on bioclast while calcite cement precipitated within the bioclast in the Middle
Bakken facies B. Plane light, BEL, 10411.3ft.

Figure 2.8: (a) The photomicrograph is showing dolomitized matrix with unreplaced calcareous bioclasts.
Plane light, BEL,10414.2ft. (b) The photomicrograph is showing zoned, planar euhedral dolomite rhomb
in the matrix of the Middle Bakken facies B. Zoning in dolomite rhomb is shown by the pink arrows. Plane
light, CP,10353.5ft.

22
Figure 2.9: The photomicrographs are showing different forms of pyrite in the Middle Bakken in Elm
Coulee Field. (a) Pyrite within dolomite rhomb is indicated by the yellow arrow, whereas pyrite around the
edge of the dolomite rhomb is shown by the white arrow. Pink haze in the matrix is due to impregnation of
the thin section by red dye and indicates microporosity. Plane light, CP,10353.41ft (b)Pyrite is replacing
calcareous fossil fragment. Plane light, BEL-10414.2ft. (c) and (d) Bioclast fragment is replaced by two
generations of cement, microcrystalline silica (white arrow) and pyrite (yellow arrow). Plane light (c) and
crossed polar (d). CP-10371.93ft.

Figure 2.10: Scanning Electron Microscope (SEM) image is showing: (a) authigenic quartz overgrowth
(yellow arrow), and (b) authigenic pore bridging illite in intercrystalline pore between the dolomite rhombs.
CP4, 10350.3ft.

23
Figure 2.11: Secondary anhydrite in different forms in the Middle Bakken in the Elm Coulee is shown in,
(a) The core photograph is showing the anhydrite pseudomorphs burrows. CP, 10354.1ft (yellow arrow).
(b) The core photograph is showing the anhydrite nodule (yellow arrow). LT, 10409.5ft. (c) and (d) The
photomicrographs are showing poikilotopic anhydrite cement in plane light and crossed polars respectively
(yellow arrow). (e) and (f) The photomicrographs are showing poikilotopic anhydrite cement in plane light
and crossed polars respectively. Note the euhedral dolomite rhombs are floating in the anhydrite cement
(yellow arrow).

24
 Petrography and Stable Isotope of the Middle Bakken Dolostones

The dolomitization process in the Middle Bakken has selectively replaced the matrix; however, the
framework components (bioclasts) has remained unaffected as shown in Figure 2.8a. The matrix of the
Middle Bakken dolostones predominantly consists of planar-subhedral to planar-euhedral dolomite rhombs,
which vary from 10 to 150µm in size. The size of the dolomite rhombs increases with the dolomite content.
Zoning is also observed in many dolomite rhombs as shown in Figure 2.8b. It was also observed that finely
crystalline dolomite rhombs appear to be floating within a poikilotopic anhydrite cement (Figure 2.11 c-f).

Figure 2.12: δ18O versus δ13C crossplot comparing the completely dolomitized samples which were
collected in this study from the Middle Bakken in Elm Coulee Field, with the reported values in the
literature. Blue and green rectangles are stable isotope values of carbon and oxygen obtained from the Late
Devonian and Early Mississippian unaltered brachiopods respectively from Mii et al.(1999) and Van
Geldern et al. (2006). Gray shaded rectangle indicates expected stable isotope values in dolomite which is
in equilibrium with the Late Devonian sea.

Stable isotope analysis of oxygen and carbon was performed to further understand the
dolomitization model of the Middle Bakken in the Elm Coulee Field. This analysis was selectively
performed only on those samples that consisted only of dolomite with no calcite impurity. XRD mineralogy
data was considered to determine the dolomite and calcite content of these samples. Middle Bakken
dolostones with more than 45 wt.% of dolomite content and dolomite index (DI) value of 1 (indicative of

25
complete dolomitization) have a mean δ18O value of -2.40‰ and δ13C of 2.07‰. The δ18O values have a
range of -2.03‰ to -2.39‰ and a standard deviation of 0.20‰. The δ13C values have a range of 1.01‰ to
2.98‰ and a standard deviation of 0.54‰. The oxygen isotope compositions of unaltered marine calcite
from Late Devonian and Early Mississippian are well constrained to a range of about -4‰ to -5.5‰ relative
to the PDB standard (Mii et al., 1999; van Geldern et al., 2006). Moreover, dolomite that originates from
Late Devonian-Early Mississippian seawater is expected to have slightly heavier δ18O values ranging from
-1.5‰ to -3‰ PDB because oxygen isotope fractionation between calcite and coevally precipitated
dolomite is about 2.5 (Major et al., 1992; Rott and Qing, 2013). Therefore, the oxygen isotope data from
Middle Bakken dolostones in the Elm Coulee Field matches with the expected oxygen isotope data in
dolomites that originated from Late Devonian-Early Mississippian sea water as shown in Figure 2.12.
Hence, the stable isotope data of oxygen indicate that the dolomitizing fluid of the Middle Bakken in the
Elm Coulee Field was derived from the seawater which was probably slightly modified.

Porosity Characterization

The total porosity of the Middle Bakken in the Elm Coulee Field, which was measured by RCA,
varies from approximately 2-10%. As shown in Figure 2.13, porosity varies considerably both within the
facies and among the facies. Within an individual location in the field, such as the cored interval of the
Coyote Putnam well (Figure 2.5), the porosity varies considerably along depth and between different facies
of the Middle Bakken. Pore architecture in the samples was studied using thin section petrography through
optical and epifluorescence microscopes and by observing whole rock samples under SEM. From thin
section petrography, intercrystalline pores and intergranular pores were identified in few samples where the
RCA derived porosity was between 7-10% and dolomite content was between 45-60 wt.%. As shown in
Figure 2.14 a & b, these intercrystalline pores are 10-15 µm in size. These visible pores, which are large
enough to be identified in thin sections through optical microscope, make up around 1% of the total porosity
of the samples. However, for samples with RCA measured porosity between 4-6% and dolomite content
less than 40wt.%, thin sections study under an epifluorescence microscope indicated the presence of
abundant sub-micron sized pores within the matrix as shown in Figure 2.14 c & d. Similar sub-micron sized
pores were identified under epifluorescence microscope by Alexandre (2011). Furthermore, in samples with
less than 4% porosity, microporosity was not observed under an epifluorescence microscope (Figure 2.14
e & f). However, it should be noted that the only indication of sub-micron sized pores under epifluorescence
microscopes is a bright orange to yellow haze. Detailed pore geometry and architecture cannot be resolved
at this magnification scale on an optical microscope. Therefore, SEM studies and MIP analysis were
performed on samples of the Middle Bakken as they consisted of fine-grained dolostones with sub-micron
sized pores as the dominant pore type.

26
 Intercrystalline pores between dolomite rhombs were identified from SEM study as shown through
the example of SEM photomicrograph in Figure 2.15a. However, most of the intercrystalline pores have
pore bridging and pore filling authigenic illite, which results in the formation of a complex pore network
(Figure 2.15b-f) and results in an overall decrease of the pore throat size. Patchy calcite cementation along
with recrystallized remnant calcite and poikilotopic anhydrite cement in the Middle Bakken in the Elm
Coulee Field also resulted in the occlusion of pore space and thereby a reduction in porosity (Figure 2.7a
and b) and Figure 2.11c-f).

Figure 2.13: Box plot of routine core analysis (RCA) derived porosity of the Middle Bakken in the Elm
Coulee Field, showing the porosity variation observed in the field. The median porosity of each of the facies
is shown by a white line within the box plots. Data consists of samples from all the facies from eight cores
whose locations are shown in Figure 2.3.

27
Figure 2.14: The photomicrographs are showing textural features corresponding to different dolomite
content in the Middle Bakken facies B in the Elm Coulee Field. (a) Intercrystalline pores between dolomite
rhombs which are sized between10-15 µm are pointed by the yellow arrows. These pores are visible in
petrographic thin section in completely dolomitized intervals. Plane light, SJ, 9731.35ft. This sample has
RCA porosity of 7.07% and dolomite content of 50 wt.% obtained from XRD. (b) Intercrystalline pores
between dolomite rhombs which are sized between10-15 µm are shown by the yellow arrows. These pores
are visible in petrographic thin section in completely dolomitized intervals. Plane light, CP, 10356.42ft.
This sample has RCA porosity of 7.86% and dolomite content of 58 wt.% obtained from XRD. (c)
Microporosity (sub-micron sized) between dolomite rhombs in the matrix is enhanced by the red dye in
petrographic thin section. Micron sized pores are not present. Plane light, BY-10472.5ft. The sample has
RCA derived porosity of 5.98% and dolomite content of 45 wt.% obtained from XRD. (d) Orange hue is
showing the microporosity (sub-micron sized) between dolomite rhombs in the matrix. Epifluorescence,
BY-10472.5ft. This sample has RCA derived porosity of 5.98%, and XRD derived dolomite content of 45
wt.%. (e) Microporosity is not visible in thin section impregnated with red dye. Plane light, CP-10367.16
This sample has RCA porosity of 3.73% and dolomite content of 36 wt.% obtained from XRD. (f) No
visible porosity in thin section impregnated with the red epifluorescence dye. Epifluorescence, CP-
10367.16ft. Sample has RCA derived porosity of 3.73%, and XRD derived dolomite content of 36 wt.%.

28
29
Figure 2.15: The Scanning Electron Microscope (SEM) photomicrographs are showing pore architecture in
the Middle Bakken facies B in the Elm Coulee Field. (a) Intercrystalline pores between dolomite rhombs
are shown by yellow arrows. CP (2), 10356.42ft. This sample has RCA porosity of 7.86% and dolomite
content of 58 wt.% obtained by XRD. (b) Yellow arrows show intercrystalline pores between the dolomite
rhombs, which is filled by pore bridging authigenic illite. The presence of illite in between the dolomite
rhombs results in a complex pore network and a decrease in the pore throat size. CP (6), 10359.29ft. This
sample has RCA porosity of 8.85% and dolomite content of 50 wt.% obtained from XRD. (c)
Intercrystalline pore between dolomite rhombs which is shown by the yellow line is filled by pore filling
authigenic illite. CP4, 10350.3ft. The sample has RCA derived porosity of 5.93%, and XRD derived
dolomite content of 51 wt.%. (d) Intercrystalline slot pores (yellow line) between dolomite rhombs which
is filled by pore bridging authigenic illite. CP4, 10350.3ft. This sample has RCA derived porosity of 5%
and dolomite content of 51 wt.% obtained from XRD.

Relationship between dolomite content and porosity

Figure 2.16 shows the cross plot between the XRD-derived dolomite content of the samples and
the corresponding RCA derived total porosity of the eight well cores representing the Elm Coulee Field.
The data points are color-coded by Dolomite Index (DI), which is defined as the ratio between total dolomite
to total carbonate, and this indicates the degree of dolomitization the samples. The samples representing
patchy calcite cement are not included in this plot because these samples do not represent the precursor

30
limestone. Following two trends were observed in Figure 2.16: 1) For dolomite content 0 to about 60 wt.%,
porosity increases with an increase in the dolomite content, 2) For dolomite content more than about 60
wt.%, porosity decreases with an increase in the dolomite content. Figure 2.17 illustrates the effect of
dolomite content on the average pore throat size through the MIP derived pore size distribution results for
seven MBM samples, for which the mineralogical composition including the dolomite content is listed in
Table 2.3. Figure 2.17 shows modal pore throat size increases with an increase in the dolomite content from
40 to 60 wt.%; however, with further increase in the dolomite content, the modal pore throat size decreases.

Figure 2.16: Crossplot between X-ray diffraction (XRD) derived dolomite content, and routine core analysis
(RCA) derived total porosity. Samples are color-coded by dolomite index (DI), which is a measure of the
degree of dolomitization. Porosity increases as dolomite content increase up to 60 wt.%. Beyond that,
porosity starts decreasing as dolomite content increases more than 60 wt.%.

Figure 2.17: Pore throat size distribution of the Middle Bakken samples from the Elm Coulee Field
determined by Mercury Intrusion Porosimetry (MIP) is shown. Pore throat size distribution is color-coded
by X-ray diffraction (XRD) dolomite content. Porosity and mineralogy details of the samples are given in
Table 2.3.

31
Table 2.3: Measured porosity from routine core analysis and dolomite, calcite, clay content derived from
X-ray diffraction and dolomite index (DI) of the samples analyzed for mercury intrusion porosimetry (MIP)
in Figure 2.17.
Sample CP-1 BY-10 BY-9 CP-2 BY-8 FE-2 CP-3
Depth (ft) 10364.4 10472.3 10467.4 10356.2 10462.3 10508.7 10353.3
Porosity (%) 4 6 9 8 11 7 4
Dolomite (wt.%) 39 45 47 58 59 61 63
Calcite (wt.%) 7 0 0 0 0 - -
Dolomite Index 0.84 1 1 1 1 1 1
Clay (wt.%) 13 8 8 5 5 - -

Based on the trend in the dolomite content vs. porosity cross plot, MIP results and SEM
photomicrographs study, four stages of porosity evolution (Stage I to IV) were identified. Stage I is
characterized by the lowest porosity (1-4%), less than 0.1 DI, and less than 15 wt.% dolomite contents. This
is the onset phase of the dolomitization of the lime mud, and therefore, pore types are still dominated by
intercrystalline pores between the partially coalesced subhedral micrite particles as shown in Figure 2.18a.
Porosity increases in stage II to about 6% and dolomite content increases to about 45 wt.%, however, DI
lies between 0.1 and 1. In this stage, subhedral to euhedral dolomite rhombs of 10-20 µm in size starts to
form as shown in Figure 2.18b. Therefore, stage II is associated with partial dolomitization because the DI
is still less than 1 and there is an occasional presence of dolomite rhombs. The dolomite rhombs at this
stage are in direct contact with subhedral to anhedral fully coalescent microcrystalline calcite grains of the
precursor limestone as shown in Figure 2.18c. Interstices between dolomite crystals are mostly filled by
abundant residual microcrystalline calcites (Figure 2.16c). Intercrystalline micropores are fewer in number
at this stage, and they selectively occur inside the clusters of dolomite crystals. Their selective occurrence
is evident from the MIP results in Figure 2.17a, which indicate the lower frequency of the modal pore throat
in the stage II samples. Moreover, the size range of the modal pore throat in these samples is between
0.015µm and 0.2µm, and they are associated with dolomite rhombs of size 10-20µm.

Stage III is marked by complete dolomitization of calcite in precursor lime mud. In this stage,
porosity further increases to a maximum of about 10%, the dolomite content increases up to 60 wt.%, and
DI attains a value of 1. The SEM photomicrograph for Stage III in Figure 2.18d shows that there is no
remnant calcite micrite and intercrystalline pores between dolomite rhombs have increased to about 70µm
in size. At this stage, these intercrystalline pores are abundantly present as evident from the MIP results in
Figure 2.17a for the corresponding samples (CP-2 and BY-8), which suggest a higher frequency of the
modal pore throat (0.3-0.4 µm). The increase in pore throat size from Stage II to Stage III is associated with
an increase in dolomite rhomb size from 10-20µm to a maximum of 70µm. In Stage IV, overdolomitization
takes place, resulting in an increase in dolomite content above 60 wt.% and a decrease in total porosity. In

32
this stage, the dolomite rhombs, which have compromised boundaries and interlocking texture, increase
above 100µm in size (Figure 2.18 e-f). In this stage, intercrystalline pores are still present; however, their
size and frequency of occurrence decreases. This is confirmed by the MIP results, which show that the pore
throat size of samples from stage IV interval ranges between 0.05-0.2µm, which is lower in comparison to
that in Stage III (Figure 2.17b).

As shown in Figure 2.5, all four stages of porosity evolution are observed in the cored interval of
the Middle Bakken facies B of the Coyote Putnam well. The precursor lime wackestone of Stage I, with the
lowest porosity (<2%), is identified in the lowermost interval of the Middle Bakken facies in this core
(10371-10376ft). Stage II of the porosity evolution, which is associated with partial dolomitization, is
observed in the interval between 10361-10371ft. In this interval, the dolomite content has an upward
increasing trend reaching a maximum of about 40 wt.% (DI 0.1-1), while the average calcite content is
around 20 wt.%. The porosity slightly increases in this interval and has a value in between 4-6%. This is
followed by Stage III of porosity evolution, which is associated with complete dolomitization observed in
the interval between 10354-10361ft. In this stage, the dolomite content varies between 45-55 wt.% and
negligible calcite content according to the XRD measurements. The average porosity in this interval is 7.8
%, which is the highest among the entire Middle Bakken interval of this core. Stage IV of porosity
evolution, which is associated with overdolomitization, is observed in this core between 10352.5-10354.5ft.
In this interval, dolomite content increases more than 60 wt.%, while porosity decreases to about 5%.

2.6 Discussion

In this section, a plausible explanation for higher dolomite content in the Elm Coulee Field is
discussed. A dolomitization model is also proposed for the Middle Bakken dolostone in the Elm Coulee
Field based on the results of this study. Furthermore, the effect of dolomitization process in the evolution
of porosity is discussed. And finally, a conceptual model of porosity evolution in the Middle Bakken of the
Elm Coulee Field is proposed.

Dolomitization in Elm Coulee Field

More calcite-rich mud was deposited in the Elm Coulee Field and its surrounding area in
comparison to other parts of the basin. During the Late Devonian-Early Mississippian time, the Williston
Basin was located near the equator, so the climate was warm during the deposition of the Middle Bakken.
This provided a favorable condition for the deposition of carbonates in the Middle Bakken. Moreover, the
Middle Bakken is siltier on the eastern and northern side of the basin, so the Canadian Shield, which is
located in the north-east of the basin, is often cited as the source of the siliciclastic sediments (Smith and
Bustin, 1996). Therefore, the Elm Coulee Field, which is located along the south-western edge of the basin

33
parallel to the Late Devonian-Early Mississippian paleo-shoreline of the Williston Basin, lay at the farthest
end of the siliciclastic sediment source during the deposition of the Middle Bakken (O’Brien et al., 2012).
Hence, only a small portion of siliciclastic sediment could have been transported near the Elm Coulee Field
(Alexandre, 2011; O’Brien et al., 2012). The clear, warm waters in the Elm Coulee region allowed more
carbonates to form in this area than in other parts of the basin. Thus, the Elm Coulee Field and its
surrounding area had more original lime mud, which was later dolomitized.

A collective analysis of stratigraphic and petrographic observations, along with stable isotope data,
indicates that the seepage reflux dolomitization by slightly modified seawater resulted in the formation of
the Middle Bakken dolostones in the Elm Coulee Field. The Middle Bakken dolostones consist of planar
sub- to euhedral dolomite rhombs in the matrix with unreplaced calcareous bioclasts (Figure 2.8). Typically,
planar -euhedral dolomite crystals form at lower temperature (<50° C) settings, while irregular, anhedral
crystals and saddle dolomite form at higher temperatures (>50° C for anhedral crystals and 60°-150° C for
saddle dolomite) (Radke and Mathis, 1980). Also, numerous examples of the pervasive replacement of the
calcitic matrix by very finely crystalline, planar-subhedral type dolomite (4 to 20 mm) have been reported
in the literature (Gregg and Sibley, 1984; Sibley and Gregg, 1987; Gregg and Shelton, 1990; Al-Aasm and
Packard, 2000). The presence of these finely crystalline dolomite rhombs is often attributed to early
dolomitization that takes place soon after deposition or during shallow burial. Therefore, these textural
characteristics of the planar sub- to-euhedral dolomite rhombs indicate that most of the Middle Bakken
dolomite was formed relatively early in the diagenetic history at lower temperature settings. Moreover, as
discussed in Section 2.5.3 and shown in Figure 2.12, the δ18O values from the Middle Bakken dolostones
in the Elm Coulee Field are also consistent with the typically expected δ18O values in dolomites that
originated from Late Devonian-Early Mississippian seawater. This further supports the early origin of the
Middle Bakken dolomites in the Elm Coulee Field and also supports the idea that the dolomitizing fluid
was derived from the seawater, which was probably modified to some extent.

Generally, two models of dolomitization namely, the seepage reflux dolomitization (Adams and
Rhodes, 1960) and the mixing-zone dolomitization (Humphrey and Quinn, 1989) are commonly accepted
to explain the formation of dolomite at an early stage of diagenesis in low-temperature settings from
seawater. Alexandre, (2011) and O’Brien et al., (2012) proposed a seepage reflux model of dolomitization
for the Middle Bakken dolomites based on indirect evidence of arid paleoclimatic conditions. However, a
seepage reflux model of dolomitization for the Middle Bakken dolomites is suggested in this study based
on stratigraphic and petrographic evidence and also paleoclimatic conditions. Witzke and Heckel (1988)
suggested that during the Late Devonian-Early Mississippian time, the Williston Basin and its surrounding
area was located near the equator and was under the influence of arid climate conditions.

34
Figure 2.18: SEM images showing petrographic features representing various degrees of dolomitization
and corresponding stages of porosity evolution in the Middle Bakken of the Elm Coulee Field. (a)
Subhedral, partially coalesced micritic particles of precursor limestone, corresponding to the Stage I of
porosity evolution. Intercrystalline sub-micron sized pores between the micrite particles. BEL, 10402ft,
Dol-13 wt.%, Calc-58 wt.%, porosity-3.64%. (b) Remnant microcrystalline calcite from the precursor
micritic lime mud is present in between the euhedral dolomite rhombs in the partially dolomitized Middle
Bakken facies B (stage II of porosity evolution). Dolomite rhombs are 15-20 µm in size. Note that micritic
calcite is in direct contact with the dolomite rhombs. BY-13, 10457.6ft, Dol-36 wt.%, Calc-24 wt.%,
porosity-2.39%. (c) Subhedral to anhedral fully coalescent remnant microcrystalline calcite from precursor
lime mud is in direct contact with the 20 µm sized dolomite rhombs in the partially dolomitized Middle
Bakken facies B (stage II of porosity evolution). (d) Intercrystalline pores between the euhedral dolomite
rhombs, in the completely dolomitized interval of the Middle Bakken corresponding to the stage III of
porosity evolution, is shown by yellow arrows. Dolomite rhombs are around 30-40 µm in size. CP (2),
10356.42ft. The sample has RCA derived porosity of 7.86%, and XRD derived dolomite content of 58
wt.%. (e) Euhedral dolomite rhombs from the overdolomitized interval (stage IV of porosity evolution)
forming interlocking texture with compromised crystal boundaries between dolomite rhombs is shown by
the black arrow. Dolomite rhombs are more than 100 µm in size. BY-14, 10457.6ft, Dol-66 wt.%, porosity-
4.25%. (f) Interlocking texture with compromised crystal boundaries (pink arrow) between euhedral
dolomite rhombs in the overdolomitized interval (stage IV of porosity evolution) of the Middle Bakken is
shown by black arrows. Dolomite rhombs are more than 100 µm in size. CP-3, 10353.25ft, Dol-63 wt.%,
porosity-4.4%.

35
36
 Mixing-zone dolomitization usually takes place under humid conditions with a stable hydrologic
setting, so it can be discarded as the likely dolomitization model for the Elm Coulee dolomites. Moreover,
Alexandre (2011) suggested that since the Middle Bakken has limited fauna and ichnofacies, it was
deposited in an arid and stressed environment which was not conducive for living organisms. And, in such
arid and evaporitic condition, dolomitization of the precursor limestone commonly takes place by the
seepage reflux process (Adams and Rhodes, 1960). Usually, such dolostones are closely associated with
evaporite deposits; however, no bed-scale evaporite deposits have yet been discovered in the Bakken
Formation. Secondary anhydrite as nodules, poikilotopic cement and pseudomorphs filling burrows are
identified in both partially and completely dolomitized intervals of the Middle Bakken in the Elm Coulee
Field as shown in Figure 2.11a-f. Adams and Rhodes (1960) suggested that the presence of secondary
anhydrite in the dolomitized interval indicates that the seepage reflux is the likely process of dolomitization.
Reactive transport models also suggest that in a seepage reflux model of dolomitization, anhydrite cement
will result in porosity occlusion in the partially dolomitized interval (Jones and Xiao, 2005). Recent studies
have shown that the absence of evaporites does not discard seepage reflux as a likely process of
dolomitization. This is because mesohaline or penesaline brines, which are much below the saturation of
gypsum or anhydrite, can be generated by very little evaporation and have the potential for dolomitizing
limestone beds by penetrating downward under the influence of gravity (Simms, 1984; Whitaker and Smart,
1990; Melim and Scholle, 2002). There are also many reported examples in literature in which shallow
water carbonates have been dolomitized extensively by the reflux of mesohaline to penesaline seawater
driven by brine density and/or sea-level fluctuations without contemporaneous precipitation of evaporites
(Sun, 1994; Qing et al., 2001; Vandeginste et al., 2006, 2009; Maliva et al., 2011; Rott and Qing, 2013).
Therefore, the petrographic and geochemical evidence indicates that the Middle Bakken dolomites were
formed at an early stage of diagenesis in the marine realm by a seepage reflux of slightly modified seawater,
which formed the mesohaline to penesaline brine.

Dolomitization process and porosity evolution

This study suggests that the dolomitization process in the Elm Coulee Field has occurred by the
volume-for-volume reaction. As discussed in Section 2.6, the Middle Bakken facies show an overall
increase in total porosity, pore throat size and dolomite rhomb size as the dolomitization process proceeds
from precursor lime mud to complete dolomitization. However, total porosity and pore throat size decreases
and dolomite rhomb size increases as dolomitization proceeds further to the overdolomitization phase. The
latter negative correlation between dolomite content and porosity in the Middle Bakken of the Elm Coulee
Field suggests that dolomitization has resulted in an increase in the bulk volume. An increase in the bulk
volume during dolomitization indicates the presence of an external source of (CO3)2- ions during

37
dolomitization, which in turn suggests that dolomitization has taken place by the volume-for-volume
reaction (Equation 2). Maliva et al., (2011) suggests that a direct contact between the precursor calcite and
dolomite rhomb with no visible gap between the growing dolomite crystal and the precursor
microcrystalline calcite indicates volume-for-volume replacement of calcite to dolomite. A direct contact
between the dolomite rhomb with the precursor microcrystalline calcite is observed in the partially
dolomitized Middle Bakken samples (Figure 2.18b-c). This further substantiates that dolomitization in the
Middle Bakken primarily occurs by volume-for-volume replacement and the (CO3)2- ions required in the
process are primarily internally sourced, along with a minor external source. However, with continued
dolomitization (stage III to stage IV), when all precursor calcite is exhausted, the externally sourced (CO3)2-
ions along with Mg2+ and Ca2+ are utilized for overdolomitization.

When bulk volume remains the same, an increase in grain or crystal size results in an increase in
pore throat size as illustrated schematically in Figure 2.19. Therefore, from Stage I to Stage III, the dolomite
rhomb size increases, which results in an increase in pore throat size. But in Stage IV, the addition of
external (CO3)2- ions results in the formation of additional dolomite, which causes an increase in the size of
the dolomite rhombs and thereby an increase in bulk volume and a corresponding decrease in pore space.
The pore space between the dolomite rhombs progressively decreases as the dolomite rhombs impinge upon
each other and finally forms an interlocking texture (Figure 2.18e-f). This results in a decrease in both
porosity and pore throat size as dolomitization proceeds from Stage III to Stage IV. Therefore, this study
supports the thesis given by Lucia (2004) that the dolomitization process takes place by volume-for-volume
replacement reaction and does not create porosity, rather it can create pore space and modify pore size.

However, in the Middle Bakken, there is an overall increase in porosity from Stage I to Stage III.
The explanation for this observation is that limestones comparatively lose more porosity through burial
compaction whereas dolostones resist compaction and retain much of their porosity (Schmoker and Halley,
1982; Lucia, 2004; Ehrenberg et al., 2006). Also, it has been observed in Cretaceous and Jurassic limestones
that mud-dominated fabric becomes more compact and less porous compared to grain-dominated fabric
(Cruz, 1997; Lucia et al., 2001; Lucia, 2004). Lucia (2004) predicted that this trend continued into the
Paleozoic carbonates. This explains why an increase in porosity is observed in the Middle Bakken facies as
precursor lime mud is progressively dolomitized.

Conceptual models also supported by reactive transport models predict that seepage reflux
dolomitization results in the highest porosity in the distal path of dolomitizing fluid and the lowest porosity
proximal to the brine source caused by the process of overdolomitization (Figure 2.20) (Jones and Xiao,
2005). In the Elm Coulee Field, all the four stages of dolomitization (stage IV to stage I) are present in the
same section/location, from the top to bottom of the Middle Bakken facies B, such that porosity is low in

38
the topmost interval of the facies B, and the highest porosity is observed in the middle of the Middle Bakken
facies B as shown in Figure 2.5. The bottom of the Middle Bakken facies B has the lowest porosity, as the
precursor limestone is not dolomitized. This sequence of porosity evolution associated with the different
stages of dolomitization indicates that the top of the Middle Bakken facies B was located proximal to the
flow pathway of mesohaline brine. Also, locally in the Elm Coulee Field area, the brine mostly percolated
downward into the precursor lime mud of the Middle Bakken and thereby progressively dolomitized the
Middle Bakken.

Conceptual Model for Porosity Evolution

A conceptual model of porosity evolution in the Middle Bakken of the Elm Coulee Field is
illustrated in Figure 2.21. As discussed in Section 2.5.5, in the Middle Bakken of the Elm Coulee Field,
porosity is closely related to the dolomitization process, and it evolved in four different stages. This model
is based on the theory of a volume-for-volume replacement reaction for the dolomitization process;
therefore, the change in total porosity did not result from dolomitization, which only changes the pore space
and pore size. However, the loss in the total porosity of the resultant dolostone was caused by the pre-
dolomitization process of patchy calcite cementation, anhydrite cementation during dolomitization,
overdolomitization, the post-dolomitization process of authigenic illite precipitation and mechanical
compaction. However, at a given location, these processes take place in different proportions, which result
in the observed variation in the porosity of the Middle Bakken in the Elm Coulee Field. As shown in Figure
2.21a, at time T1, precursor calcareous sediment of the Middle Bakken consisting of silty lime mud was
deposited in an epeiric platform in a shallow water setting.

Figure 2.19: Schematic diagram illustrating an increase in the size of dolomite rhombs results in an increase
in pore size.

39
Figure 2.20: Conceptual model illustrating porosity evolution both spatially and temporally due to seepage
reflux dolomitization in a ramp setting. As seepage reflux continues from time T1 through T2 continues,
limestones are progressively dolomitized, with the formation of dolomite cement near the source of the
brine and anhydrite cement at the distal end, which in turn results in variation of porosity. (Jones and Xiao,
2005)

At time T1 the entire sediment column (d3-d0), which consisted of lime mud, was in Stage I of the
porosity evolution. As shown in Figure 2.21b, at time T2, the bioclasts present within the lime mud dissolved
and got precipitated in the deeper sediment column (d1-d0) prior to any mechanical compaction, resulting
in the formation of early patchy calcite cement (Section 2.5.2). This led to a slight reduction in the porosity
of the sediment column d0-d1 prior to dolomitization; however, the entire sediment column was still in Stage
I of the porosity evolution. This is followed by the seepage reflux of slightly modified seawater into the
Middle Bakken sediment, almost contemporaneously with the deposition of the Middle Bakken at time T3,
as shown in Figure 2.21c. The arid climatic condition during deposition of the Middle Bakken resulted in
an increased evaporation of seawater. This led to the formation of mesohaline brine, which percolated both
downward and laterally in the precursor lime mud of the Middle Bakken facies. The mesohaline fluid
preferentially flowed through the porous lime mud and avoided the impervious patchy calcite cement. As
the reflux of mesohaline brine started, the shallow sediment column in interval d3-d2 was exposed to the

40
brine, which had the highest Mg2+/Ca2+ ratio. As a result, dolomite rhombs of 10-20µm developed at the
expense of calcite in the lime mud, and along with that, some intercrystalline pores also formed between
the rhombs. Moreover, the modal pore throat size also increased at this stage as the overall grain size of the
sediments in interval d3-d2 increased due to the replacement of the smaller micritic calcite crystals (<1µm)
by larger dolomite rhombs (10-20 µm). As shown in Figure 2.21c, as the mesohaline brine percolated in
intervals deeper than d2, the Mg2+/Ca2+ ratio progressively decreased. Therefore, the mesohaline brine fluid,
which percolated below d1, or the deepest Middle Bakken sediments, had almost lost its dolomitizing
potential. This, in turn, resulted in the precipitation of the secondary anhydrite cement in sediment interval
d2-d1 causing a minor loss in total porosity. So, at time T3, the interval between d2-d3 was partially
dolomitized with porosity evolution at Stage II, while interval below d1 was in Stage I, and the interval
between d2-d1 was in transition stage (Figure 2.21c).

The process of reflux of mesohaline brine and progressive decrease in the Mg2+/Ca2+ ratio in deeper
sections of Middle Bakken continued uninterruptedly. Therefore, the shallower interval between d3-d2 was
exposed to comparatively more pore volumes of mesohaline fluid with a higher Mg2+/Ca2+ ratio in
comparison to the deeper sections. Subsequently, at time T4, the remaining calcite in the partially
dolomitized interval between d3-d2 was completely replaced by larger dolomite rhombs of 20-50 µm, and
along with that more intercrystalline pores were formed between the rhombs. The modal pore throat size in
interval d2-d3 also got enhanced in the process as the remaining micritic calcite was replaced by even larger
dolomite rhombs. Therefore, as shown in Figure 2.21d at time T4, the interval d3-d2 was completely
dolomitized and was in Stage III of porosity evolution, and d2-d1 was partially dolomitized or at Stage II,
while the interval below d1 remained largely unaltered or in Stage I. As discussed previously, the total
porosity of both completely dolomitized (d2-d3) and partially dolomitized (d1-d2) intervals remained same
as the precursor lime mud, since dolomitization does not change the total porosity. However, porosity
decreased in the partially dolomitized interval below d1 due to early patchy calcite cementation and
secondary anhydrite cementation.

At time T5, additional dolomite was precipitated in completely dolomitized interval d3-d2 as more
mesohaline fluid percolated through it due to the continued availability of Ca2+, Mg2+ and (CO3)2- ions. This
resulted in the overdolomitization of interval d3-d2 as shown in Figure 2.21e. As more dolomite was
precipitated, the size of the dolomite rhombs continued to increase up to 150µm, thereby decreasing the
porosity and the pore throat size. Even after overdolomitizing the d3-d2 interval, the mesohaline fluid still
had significant dolomitizing potential; however, the Mg2+ /Ca2+ratio decreased.

41
Figure 2.21: Cartoon illustrating the progressive evolution of porosity in the Middle Bakken of the Elm
Coulee Field from time T1 through T6, due to seepage reflux dolomitization of precursor lime mud. Different
depths in the sediment column are represented by d0 to d3. (a) At time T1, the precursor lime mud got
deposited, and the entire sediment column has an overall high porosity. (b) At time T2, early patchy calcite
cementation results in porosity loss at depths between d0 to d1. (c) At time T3, seepage reflux of mesohaline
brine has started. Sediment column from d3 to d2 is partially dolomitized, with the degree of dolomitization
decreasing down the depth. No change in total porosity is observed between lime mud and partially
dolomitized limestone due to dolomitization. However, pore space and pore size are modified as
intercrystalline pores start to form between dolomite rhombs. Moreover, the formation of anhydrite cement
may result in minor loss of porosity near depth d1. (d) At time T4, seepage reflux of mesohaline brine
continues such that the partially dolomitized lime mud between d3 to d2 is completely dolomitized with no
change in total porosity. While lime mud between d2 to d1 is partially dolomitized with the degree of
dolomitization decreasing down the depth such that below depth d0, precursor lime mud remains unaltered.
There is no change in total porosity due to dolomitization. However, pore space and pore size are modified
as intercrystalline pores forms between dolomite rhombs. Moreover, the formation of anhydrite cement
may result in minor loss of porosity between depth d1 to d0. (e) At time T5, seepage reflux of mesohaline
brine continues such that the completely dolomitized interval between d3 to d2 is overdolomitized with a
loss in porosity due to the formation of dolomite cement. While partially dolomitized lime mud between d2
to d1 is completely dolomitized. Below depth d1, the precursor lime mud is partially dolomitized with the
degree of dolomitization decreasing down the depth such that below depth d0, precursor lime mud remains
unaltered. There is no change in total porosity due to dolomitization. However, pore space and pore size
are modified as intercrystalline pores forms between dolomite rhombs. (f) At time T5, after cessation of the
seepage reflux dolomitization, mechanical compaction of the sediment column occurs due to burial. This
results in differential compaction of the sediment column such that the precursor lime mud is more
compacted and therefore, loses more porosity in comparison to the dolostones.

42
43
 Continued downward percolation of this mesohaline fluid with reduced Mg2+/Ca2+ ratio in the
deeper section resulted in complete dolomitization of the previously partially dolomitized intervals d2-d1.
The precursor lime mud in interval d1-d0 was partially dolomitized by this fluid, while the precursor lime
mud below d0 remained unaltered. Therefore, as shown in Figure 2.21e at time T5, the interval d3-d2 was
overdolomitized and was in Stage IV of porosity evolution, and d2-d1 was completely dolomitized and was
in Stage III of porosity evolution. The interval d1-d0 was partially dolomitized, while the interval below d0
remained largely unaltered or in Stage-I.

Finally, mechanical compaction due to burial started after the above-discussed dolomitization
process was complete. As discussed in Section 2.2, dolostones can endure compaction more effectively
than limestones. Therefore, porosity loss due to mechanical compaction in limestones is greater than the
corresponding dolostones. Hence, as shown in Figure 2.21f., porosity loss due to compaction in the Middle
Bakken in the Elm Coulee Field was in the following order: below d0>d0-d1>d1-d2>d2-d3. After
dolomitization, authigenic illite clay was precipitated within the intercrystalline pore space, which further
decreased both the total porosity and pore throat size. Thus, porosity evolution in the Middle Bakken of
Elm Coulee Field was dependent on three diagenetic factors: the degree of dolomitization, burial
compaction, and authigenic illite precipitation.

2.7 Conclusions

Based on this study, the following conclusions are drawn:

1. Integration of stratigraphic and petrographic observations along with the analysis of stable isotope
data of oxygen and carbon indicates that precursor silty lime mud in the Middle Bakken in the Elm
Coulee Field was dolomitized at varying degrees by seepage-reflux of mesohaline to penesaline
brine. The seepage-reflux dolomitization of the Middle Bakken took place soon after the deposition
of the lime mud.
2. The Middle Bakken consists of sub-micron sized intercrystalline pores between dolomite rhombs.
The presence of pore bridging and pore lining authigenic illite clay within these intercrystalline
pores results in a complex pore network thereby reducing the pore throat size.
3. Porosity evolution in the Middle Bakken silty dolostones of the Elm Coulee Field is closely
associated with the degree of dolomitization. Four stages of porosity evolution (stage I to IV) were
identified in the Middle Bakken based on the dolomite content, dolomite index (DI) and other
textural features. Porosity evolution in Stage I is associated with the precursor lime mud (DI is less
than 0.1, and the dolomite content is less than 15 wt.%). Porosity evolution in Stage II is associated
with the partial dolomitization of the precursor lime mud (DI varies between 0.1 to 1 and dolomite
content is less than 45 wt.%). Porosity evolution in Stage III is associated with the complete

44
 dolomitization of the precursor lime mud such that DI=1 and the dolomite content is less than 60
wt.%. Porosity evolution in Stage IV is associated with overdolomitization of the precursor lime
mud (DI=1 and dolomite content is more than 60 wt.%). Total porosity and pore throat size of the
Middle Bakken increases from Stage I to III as the degree of dolomitization increases from the
initial stages of dolomitization in the precursor lime mud to complete dolomitization. However,
both total porosity and pore throat size start decreasing from Stage III to Stage IV as the degree of
dolomitization increases further from complete dolomitization to overdolomitization.
4. Textural features of the Middle Bakken dolostones indicate that dolomitization has taken place by
the volume-for-volume reaction such that the dolomitizing mesohaline fluid supplied (CO3)2- ions
along with the Mg2+ ions. So, dolomitization by replacement of calcite did not create porosity but
modified the pore space and pore throat size in the Middle Bakken. However, overdolomitization
resulted in a decrease in porosity due to an increase in the crystal size of the dolomite rhombs.
5. The total porosity of the completely dolomitized interval (Stage III) of the Middle Bakken is greater
compared to the precursor lime mud (Stage I) because limestone loses more porosity due to burial
compaction compared to dolostones.
6. Early patchy calcite cement, anhydrite cement, and authigenic illite clay are other diagenetic factors
that reduce the porosity of the Middle Bakken.
7. A conceptual model of textual and porosity evolution in the Middle Bakken in the Elm Coulee
Field is proposed. According to this model, the shallower intervals of the Middle Bakken were
subjected to more pore volumes of the Mg-rich mesohaline brine fluid, and therefore have been
overdolomitized. However, as the mesohaline fluid percolated downward into deeper intervals, the
Mg/Ca ratio progressively decreased which resulted in gradually less dolomitized intervals.

2.8 References

Adams, J. E., and M. L. Rhodes, 1960, Dolomitization by seepage refluxion: AAPG Bulletin, v. 44, no. 12,
p. 1912–1920.
Al-Aasm, I. S., and J. J. Packard, 2000, Stabilization of early-formed dolomite: a tale of divergence from
two Mississippian dolomites: Sedimentary Geology, v. 131, no. 3–4, p. 97–108.
Al-Awadi, M., W. J. Clark William Ray Moore, M. Herron Tuanfeng Zhang Weishu Zhao, N. Hurley
Dhahran, S. Arabia Djisan Kho East Ahmadi, K. Bernard Montaron Dubai, U. Fadhil Sadooni, and T.
Smithson, 2009, Dolomite: Perspectives on a Perplexing Mineral: Oilfield Review, Autumn, v. 121,
no. 3.
Alexandre, C. S., 2011, Reservoir characterization and petrology of the Bakken Formation, Elm Coulee
Field, Richland County, MT: MS thesis, Colorado School of Mines, 175 p.
Allan, J. R., and W. D. Wiggins, 1993, Dolomite Reservoirs: American Association of Petroleum
Geologists, no. 26.

45
Angulo, S., and L. A. Buatois, 2012, Integrating depositional models, ichnology, and sequence stratigraphy
in reservoir characterization: The middle member of the Devonian--Carboniferous Bakken Formation
of subsurface southeastern Saskatchewan revisited: AAPG bulletin, v. 96, no. 6, p. 1017–1043.
Brennan, S. W., 2016, Integrated characterization of Middle Bakken diagenesis, Williston Basin, North
Dakota, USA: MS Thesis, Colorado School of Mines.
Cantrell, D., P. Swart, and R. Hagerty, 2004, Genesis and characterization of dolomite, Arab-D reservoir,
Ghawar field, Saudi Arabia: GeoArabia, v. 9, no. 2, p. 11–36.
Choquette, P. W., and E. E. Hiatt, 2008, Shallow-burial dolomite cement: A major component of many
ancient sucrosic dolomites: Sedimentology, v. 55, no. 2, p. 423–460.
Cruz, W. M., 1997, Study of Albian carbonate analogs: Cedar Park Quarry, Texas, USA, and Santos Basin
Reservoir, Southeast Offshore of Brazil.: MS Thesis, The University of Texas at Austin,
Davies, G. R., 1979, Dolomite reservoir rocks: Processes, controls, porosity development, in C. H. Moore,
ed., Geology of carbonate porosity: AAPG Continuing Education Course Note Series 11, p. C1–C17.
Deville de Periere, M., C. Durlet, E. Vennin, L. Lambert, R. Bourillot, B. Caline, and E. Poli, 2011,
Morphometry of micrite particles in cretaceous microporous limestones of the middle east: Influence
on reservoir properties: Marine and Petroleum Geology, v. 28, no. 9, p. 1727–1750.
Ehrenberg, S. N., G. P. Eberli, M. Keramati, and S. A. Moallemi, 2006, Porosity-permeability relationships
in interlayered limestone-dolostone reservoirs: AAPG Bulletin, v. 90, no. 1, p. 91–114.
Ehrenberg, S. N., O. Walderhaug, and K. Bjerlykke, 2012, Carbonate porosity creation by mesogenetic
dissolution: Reality or illusion? AAPG Bulletin, v. 96, no. 2, p. 217–225.
van Geldern, R., M. M. Joachimski, J. Day, U. Jansen, F. Alvarez, E. A. Yolkin, and X. P. Ma, 2006,
Carbon, oxygen and strontium isotope records of Devonian brachiopod shell calcite: Palaeogeography,
Palaeoclimatology, Palaeoecology, v. 240, no. 1–2, p. 47–67.
Gent, V. A., 2011, Fourier Analysis of the Laminated Facies of the Middle Bakken Member, Sanish Parshall
Field, Mountrail County, North Dakota: MS thesis, Colorado School of Mines, 135 p.
Gregg, J. M., and D. F. Sibley, 1984, Epigenetic dolomitization and the origin of xenotopic dolomite
texture: Journal of Sedimentary Petrology, v. 54, p. 908 – 931.
Halley, R. B., and J. W. Schmoker, 1983, High-porosity Cenozoic carbonate rocks of south Florida:
progressive loss of porosity with depth.: AAPG Bulletin, v. 67, no. 2, p. 191–200.
Humphrey, J. D., and T. M. Quinn, 1989, Coastal mixing zone dolomite, forward modeling, and massive
dolomitization of platform-margin carbonates: Journal of Sedimentary Research, v. 59, no. 3, p. 438–
454.
Jones, G. D., and Y. Xiao, 2005, Dolomitization, anhydrite cementation, and porosity evolution in a reflux
system: Insights from reactive transport models: AAPG Bulletin, v. 89, no. 5, p. 577–601.
Kaczmarek, S. E., S. M. Fullmer, and F. J. Hasiuk, 2015, A universal classification scheme for the
microcrystals that host limestone microporosity: Journal of Sedimentary Research, v. 85, no. 10, p.
1197–1212.
Kent, D. M., and J. E. Christopher, 1994, Geological history of the Williston Basin and Sweetgrass arch:
Geological atlas of the Western Canada Sedimentary Basin: Canadian Society of Petroleum Geologists
and Alberta Research Council, p. 421–429.
Land, L. S., 1983, The application of stable isotope studies of the origin of dolomite and to problems of
diagenesis of clastic sediments: in M. A. Arthur, ed., Stable isotopes in sedimentary geology, p. 1–22.

46
Landes, K. K., and A. Arbor, 1946, Porosity through dolomitization: American Association of Petroleum
Geologist Bulletin, v. 30, no. 3.
Gerhard, L. C., S. B. Anderson, and D. W. Fischer, 1990, Petroleum geology of the Williston Basin: AAPG
Memoir, v. 51, p. 507-559.
LeFever, J. A., C. D. Martiniuk, E. F. R. Dancsok, and P. A. Mahnic, 1991, Petroleum potential of the
middle member, Bakken Formation, Williston Basin: in J. E. Christopher and F. Haidl, eds.,
Proceedings of the Sixth International Williston Basin Symposium: Saskatchewan Geological Society,
Special publication 11, p. 74-94.
Longman, M. W., T. G. Fertal, and J. S. Glennie, 1983, Origin and geometry of Red River dolomite
reservoirs, western Williston Basin.: American Association of Petroleum Geologists Bulletin, v. 67,
no. 5, p. 744–771.
Loucks, R. G., R. M. Reed, S. C. Ruppel, and U. Hammes, 2012, Spectrum of pore types and networks in
mudrocks and a descriptive classification for matrix-related mudrock pores: AAPG Bulletin, v. 96, no.
6, p. 1071–1098.
Loucks, R. G., and M. Ulrich, 2015, Origin and Characterization of the Nanopore / Micropore Network in
the Leonardian Clear Fork Reservoirs in the Goldsmith Field in Ector Co., Texas: In AAPG Annual
Convention and Exhibition, no. 51164.
Lucia, F. J., 2004, Origin and petrophysics of dolostone pore space: Geological Society, London, Special
Publications, v. 235, no. 1, p. 141–155.
Lucia, F. J., 1995, Rock-Fabric / Petrophysical Classification of Carbonate Pore Space for Reservoir
Characterization 1: v. 9, no. 9, p. 1275–1300.
Lucia, F. J., J. W. Jennings Jr., M. Rahnis, and F. O. Meyer, 2001, Permeability and rock fabric from
wireline logs, Arab-D reservoir, Ghawar field, Saudi Arabia.: GeoArabia, v. 6, no. 4, p. 619–646.
Lucia, F. J., and R. G. Loucks, 2013, Micropores Development in Carbonate Mud: Early Development and
Petrophysics: Gulf Coast Association of Geological Societies Journal, v. 2, p. 1–10.
Lucia, F. J., and R. P. Major, 1994, Porosity evolution through hypersaline reflux dolomitization: Wiley
Online Library, p. 325–341.
Major, R. P., R. M. Lloyd, and F. J. Lucia, 1992, Oxygen isotope composition of Holocene dolomite formed
in a humid hypersaline setting: Geology, v. 20, no. 7, p. 586–588.
Maliva, R. G., D. A. Budd, E. A. Clayton, T. M. Missimer, and J. A. D. Dickson, 2011, Insights Into the
Dolomitization Process and Porosity Modification in Sucrosic Dolostones, Avon Park Formation
(Middle Eocene), East-Central Florida, U.S.A.: Journal of Sedimentary Research, v. 81, no. 3, p. 218–
232.
Melim, L. A., and P. A. Scholle, 2002, Dolomitization of the Capitan Formation forereef facies (Permian,
west Texas and New Mexico): seepage reflux revisited: Sedimentology, v. 49, p. 1207–1227.
Mii, H. S., E. L. Grossman, and T. E. Yancey, 1999, Carboniferous isotope stratigraphies of North America:
Implications for Carboniferous paleoceanography and Mississippian glaciation: Bulletin of the
Geological Society of America, v. 111, no. 7, p. 960–973.
Morrow, D. W., 1982, Dolomite-Part 1: The Chemistry of Dolomitization and Dolomite Precipitation:
Geoscience Canada, v. 9, no. 1, p. 5–13.
O’Brien, D., R. Larson Jr., R. Parham, B. Thingelstad, W. Aud, R. Burns, and L. Weijers, 2012, Using
Real-Time Downhole Microseismic To Evaluate Fracture Geometry for Horizontal Packer-Sleeve

47
 Completions in the Bakken Formation, Elm Coulee Field, Montana: SPE Production & Operations,
no. 1, p. 27–43.
Pitman, J. K., L. C. Price, and J. a LeFever, 2001, Diagenesis and Fracture Development in the Bakken
Formation, Williston Basin: Implications for Reservoir Quality in the Middle Member: USGS
Professional Paper 1653, p. 19.
Pittman, E. D., 1971, Microporosity in carbonate rocks: AAPG Bulletin, v.55, no. 10, p. 1873–1881.
Powers, R. W., 1962, Arabian Upper Jurassic Carbonate Reservoir: in W.E. Ham, ed., Classification of
Carbonate Rocks, AAPG Memoir, v. 38, p. 122–192.
Pramudito, A., 2008, Depositional facies, diagenesis, and petrophysical analysis of the Bakken Formation,
Elm Coulee Field, Williston Basin, Montana: MS Thesis, Colorado School of Mines. 194 p.
Price, L. C., T. Ging, T. Daws, A. Love, M. Pawlewicz, and D. Anders, 1984, Organic Metamorphism in
the Mississippian-Devonian Bakken Shale, North Dakota Portion of the Williston Basin: in J.
Woodward, F. F. Meissner and J. L. Clayton, eds., Hydrocarbon Source Rocks of the Greater Rocky
Mountain Region, Rocky Mountain Association of Geologists, p. 83–133.
Purser, B. H., A. Brown, and D. M. Aissaoui, 1994, Nature, origin and evolution of porosity in dolomites:
Dolomites: A Volume in Honour of Dolomieu. B. Purser, M. Tucker and D. Zenger (eds.).
International Association of Sedimentologists, Special Publication, no. 21, p. 283–308.
Qing, H., D. W. J. Bosence, and E. P. F. Rose, 2001, Dolomitization by penesaline sea water in early
Jurassic peritidal platform carbonates, Gibraltar, western Mediterranean: Sedimentology, v. 48, no. 1,
p. 153–163.
R. C. Murray, 1960, Origin of Porosity in Carbonate Rocks: SEPM Journal of Sedimentary Research, v.
Vol. 30, no. 1, p. 59–84.
Radke, B. M., and R. L. Mathis, 1980, On the Formation and Occurrence of Saddle Dolomite: Journal of
Sedimentary Petrology, v. 50, no. 4, p. 1149–1168.
Rolfs, S. A., 2015, Integrated geomechanical, geophysical, and geochemical analysis of the Bakken
Formation, Elm Coulee field, Williston Basin, Montana: MS thesis, Colorado School of Mines, 122
p.
Rott, C. M., and H. Qing, 2013, Early Dolomitization and Recrystallization in shallow marine carbonates,
Mississippian Alida Beds, Williston Basin (Canada): Evidence from Petrography and Isotope
Geochemistry: Journal of Sedimentary Research, v. 83, p. 928–941.
Saller, A. H., and N. Henderson, 1998, Distribution of porosity and permeability in platform dolomites:
insight from the Permian of west Texas: AAPG Bulletin, v. 82, no. 8, p. 1528–1550.
Saller, A. H., S. Walden, S. Robertson, R. Nims, J. Schwab, H. Hagiwara, and S. Mizohata, 2004, Three-
Dimensional Seismic Imaging and Reservoir Modeling of an Upper Paleozoic “Reefal” Buildup,
Reinecke Field, West Texas, United States: Seismic imaging of carbonate reservoirs and systems:
AAPG Memoir 81, p. 107–122.
Schmoker, J. W., and R. B. Halley, 1982, Carbonate Porosity Versus Depth: A Predictable Relation for
South Florida: AAPG Bulletin, v. 66, no. 12, p. 2561–2570.
Schmoker, J. W., K. B. Krystinik, and R. B. Halley, 1985, Selected Characteristics of Limestone and
Dolomite Reservoirs in the United States: AAPG Bulletin, v. 69, no. 5, p. 733–741.
Sibley, D. F., and Gregg J.M., 1987, Classification of dolomite rock textures.: Journal of Sedimentary
Petrology, v. 57, no. 6, p. 967–975.

48
Simenson, A., 2010, Depositional facies and petrophysical analysis of the Bakken Formation, Parshall
Field, Mountrail County, North Dakota: MS Thesis, Colorado School of Mines, 219 p.
Simms, M., 1984, Dolomitization by Groundwater-flow Systems in Carbonate Platforms: Transactions of
the Gulf Coast Association of Geological Sciences, v. 24, p. 411–420.
Smith, M. G., and R. M. Bustin, 1996, Lithofacies and paleoenvironments of the Late Devonian and Early
Mississippian Bakken Formation, Williston Basin: Bulletin of Canadian Petroleum Geology, v. 44,
no. 3, p. 495–507.
Smith, M. G., and R. M. Bustin, 1998, Production and preservation of organic matter during deposition of
the Bakken Formation (Late Devonian and Early Mississippian), Williston Basin: Palaeogeography,
Palaeoclimatology, Palaeoecology, v. 142, no. 3–4, p. 185–200.
Sonnenberg, S. A., and A. Pramudito, 2009, Petroleum geology of the giant Elm Coulee field, Williston
Basin: AAPG Bulletin, v. 93, no. 9, p. 1127–1153.
Sun, S. Q., 1994, A reappraisal of dolomite abundance and occurrence in the Phanerozoic: Journal of
Sedimentary Research, v. 64, no. 2a, p. 396–404.
Sun, S. Q., 1995, Dolomite Reservoirs: Porosity Evolution and Reservoir Characteristics 1: AAPG Bulletin,
v. 2, no. 2, p. 186–204.
Taylor, A. M., and R. Goldring, 1993, Description and analysis of bioturbation and ichnofabric: Journal of
the Geological Society, v. 150, no. 1, p. 141–148.
Tucker, M. E., and V. P. Wright, 1990, Carbonate Sedimentology: Oxford, UK, Blackwell Publishing Ltd.
Vandeginste, V., R. Swennen, S. A. Gleeson, R. M. Ellam, K. Osadetz, and F. Roure, 2006, Development
of secondary porosity in the Fairholme carbonate complex (southwest Alberta, Canada): Journal of
Geochemical Exploration, v. 89, no. 1–3, p. 394–397.
Vandeginste, V., R. Swennen, M. H. Reed, R. M. Ellam, K. Osadetz, and F. Roure, 2009, Host rock
dolomitization and secondary porosity development in the Upper Devonian Cairn Formation of the
Fairholme carbonate complex (South-west Alberta, Canadian Rockies): Diagenesis and geochemical
modelling: Sedimentology, v. 56, no. 7, p. 2044–2060.
Wang, G., P. Li, F. Hao, H. Zou, and X. Yu, 2015, Dolomitization process and its implications for porosity
development in dolostones: A case study from the Lower Triassic Feixianguan Formation, Jiannan
area, Eastern Sichuan Basin, China: Journal of Petroleum Science and Engineering, v. 131, p. 184–
199.
Warren, J., 2000, Dolomite: occurrence, evolution and economically important associations: Earth-Science
Reviews, v. 52, no. 1–3, p. 1–81.
Webster, R. L., 1984, Petroleum source rocks and stratigraphy of the Bakken Formation in North Dakota:
in Woodward J., F. F. Meissner, J. L. Clayton, eds., Hydrocarbon source rocks of the greater Rocky
Mountain region, Rocky Mountain Association of Geologists, p. 57 -82.
Weyl, P, K., 1960, Porosity through dolomitization conservation of mass requirements: Journal of
Sedimentary Petrology, v. 30, p. 85-90.
Whitaker, F. F., and P. L. Smart, 1990, Active circulation of saline ground waters in carbonate platforms:
Evidence from the Great Bahama Bank: Geology, v. 18, p. 200–203.
Witzke, B. J., and P. H. Heckel, 1988, Paleoclimatic indicators and inferred Devonian paleolatitudes of
Euramerica: In: N.J. McMillan, A.F. Embry, D.J. Glass, eds., Devonian of the World, I. Canadian
Society of Petroleum Geologists, Memoir, vol. 14, pp. 49 – 63.

49
 3. CHAPTER 3

FACIES CHARACTERIZATION AND CHEMOSTRATIGRAPHY OF ORGANIC-

RICH UPPER BAKKEN SHALE, WILLISTON BASIN

Recent exploration success and historical production from the Upper Bakken Shale (UBS) highlight
its future resource potential. This emphasizes the need for comprehensive characterization of the UBS with
respect to its facies and internal stratigraphic framework. This paper integrates results of thin-section
petrography with the organic and inorganic geochemical analyses to characterize the facies, and to identify
the laterally correlatable chemostratigraphic units of the UBS. Using these findings, this paper also
addresses various gaps and inconsistencies in the current understanding of the prevailing sedimentary
processes and depositional conditions for the organic-rich UBS.

In this study, handheld X-Ray Fluorescence (XRF) analyzer was used for elemental concentration
data acquisition of sixteen well cores at 6-inch intervals. Thirteen well cores were described, and thin
sections from five wells were examined. The cores used in this study were spread over both basin margin
and basin depocenter areas. Six lithofacies, F1 through F6, were identified in the UBS from core
descriptions and thin-section petrographic studies. 95% of the UBS consists of three main facies: siliceous
mudstone (F1), massive to finely laminated silt-bearing mudstone (F2) and, macrofossil-bearing silt-rich
mudstone (F4). Based on the trends of major and trace element concentrations obtained from the XRF data,
three regionally correlatable chemostratigraphic packages (sub-units 1a, 1b, and Unit-2), were identified in
the UBS. The elemental concentration data were also used as proxies to understand lateral and temporal
variations in the detrital and biogenic sediment influx and to determine the paleoredox conditions of the
basin during deposition of UBS.

Based on the results of this study it was proposed that influx of silt-size detrital sediments have
multiple sources, especially during the deposition of Unit-2. Along with eolian silt from the northeast, some
detrital silt was derived from the southern basin margin. Redox conditions varied both temporally and
laterally during the deposition of the UBS. Sub-units 1a and 1b were deposited in a persistently euxinic
condition in most locations across the basin. Unit-2 was deposited in a much less reducing condition, which
varied from sub-oxic to intermittently euxinic. The maximum flooding surface was located in sub-unit 1b,
below the Unit-2 and sub-unit 1b contact.

3.1 Introduction

The organic-rich Upper Bakken Shale (UBS) of the Bakken Formation is a world-class source rock
with an average total organic content (TOC) of 11 wt.% (LeFever et al., 1991; Sonnenberg and Pramudito,

50
2009; Jin, 2014). Although the dolomitic siltstone of the Middle Bakken reservoir is the primary target of
current production activities in the prolific Bakken Play, the naturally fractured UBS was a resource play
from 1976 to 2000 (Sonnenberg, 2014). The historical production and drilling activities in the UBS were
primarily based in the Billings Nose structure, around the Bakken Fairway area (Sperr, 1990; Carlisle et al.,
1992). Some recent exploration efforts in the UBS in the southwest flank area of the Elm Coulee Field,
Montana, have yielded favorable results, which substantiates the resource potential of the UBS
(Sonnenberg, 2014). This emphasizes the necessity of developing a basin-wide internal stratigraphic
framework for the UBS for the identification of the target intervals and areas to aid future exploration and
production efforts. The UBS, like other organic-rich shales and mudrocks, are particularly difficult to
characterize, due to their macroscopic homogeneity and mineralogic complexity (Sano et al., 2013; Tinnin
and Darmaoen, 2016). Complementing traditional analytical petrographic techniques with new methods,
such as chemostratigraphic analysis, is expected to provide a better understanding of the UBS, with respect
to identifying its regionally correlated sub-units and their depositional conditions.

Previous studies on the UBS describe it as a massively to finely laminated, fissile, siliceous, dark
black, organic-rich shale with abundant pyrite content (Webster, 1984; Hayes, 1985; Smith and Bustin,
1996, 1998), but few studies in the past have focused on the possibility of lateral and temporal variations
in the UBS with respect to its major lithofacies and their lateral and vertical distributions. For example,
Egenhoff and Fishman (2013) identified three major facies in the UBS based on petrographic studies. The
authors suggested that the UBS was deposited in a predominantly dysoxic setting based on the observation
of extensive vertical microburrows and sedimentary structures indicating bedload transport processes
observed in petrographic thin-sections. This interpretation, however, was later contradicted in a study by
Schieber (2014), which suggested that the features in petrographic thin sections are essentially preparation
artifacts. This highlights the need for integrating petrographic analysis with chemostratigraphic methods
for characterizing organic-rich shales and their depositional conditions. Chemostratigraphic studies are
based on inorganic geochemical results, such as elemental concentrations, which are more sensitive to the
paleoenvironment and paleoredox conditions during their deposition (Sano et al., 2013; Nakamura, 2015;
Nance and Rowe, 2015; Turner et al., 2016). Inorganic geochemical characterization helps define the
chemostratigraphic packages in shales based on the abundancies of different elements, which in turn helps
define a robust correlation framework in shales (Sano et al., 2013; Ramkumar, 2015). Moreover, the recent
advancement in handheld energy dispersive X-ray fluorescence (ED-XRF) analysis provides a non-
destructive method for the collection of high-resolution chemostratigraphic datasets directly from the cores
(Mainali, 2011; Kocman, 2014; Nakamura, 2015). Kocman (2014) performed chemostratigraphic analysis
of the UBS for seven cores along the eastern part of the Williston Basin, and thereby identified two laterally
correlatable chemostratigraphic units of UBS; however, the significance of these chemostratigraphic units,

51
with respect to variations in lithofacies and depositional conditions for the UBS, was not discussed by the
author.

In this study, I integrate core descriptions, thin-section petrography, chemostratigraphic analysis,

and TOC results from multiple wells across the basin to accomplish the following objectives: i) identify
lithofacies from core descriptions and thin-section petrography and interpret the sedimentary processes; ii)
identify laterally correlatable chemostratigraphic units and integrate the lithofacies with inorganic
geochemical data used as elemental proxies to understand the depositional condition of each unit; iii)
determine the areal extent and distribution of these units in terms of their thickness; and iv) infer the
stratigraphic position of the maximum flooding surface in the UBS.

3.2 Geological Setting

The Bakken Formation, which covers portions of western North Dakota and northeastern Montana
in the U.S. and southern Saskatchewan and southwestern Manitoba in Canada, were deposited in the
Williston Basin during the Late Devonian-Early Mississippian time, as seen in Figure 3.1 (LeFever et al.,
1991; Smith and Bustin, 1996). The Williston Basin is an elliptical intracratonic basin located along the
western edge of the Canadian Shield, and it extends over an area of around 100,000 square miles (Kent and
Christopher, 1994). The basin is bordered on the east and north-east by the Sioux Arch and bounded on the
north by the Punnichy Arch fronting the Saskatchewan Monocline. The exposed Precambrian Canadian
Shield is commonly considered as the eastern and northern edge of the basin. The Sweetgrass Arch of
northern Montana and southeastern Alberta limits the western extent of the basin, whereas the Black Hills
Uplift and Miles City Arch mark the southern boundary of the basin (Gerhard et al., 1982; Kent and
Christopher, 1994). In the Canadian region, the northeast trending Sweet Grass Arch separates the Williston
Basin from the Alberta Basin, which is a northwest-trending trough in front of the Cordilleran Fold and
Thrust Belt (Wright et al., 1994). As shown in Figure 3.2, a continuous sediment column of about 16,000
feet accumulated in the Williston Basin in between Cambrian and Tertiary ages (Gerhard et al., 1990;
LeFever et al., 1991). Cyclical transgressions and regressions characterize the sedimentation in the basin,
which has repeated deposition of carbonates and siliciclastics with several unconformities in between.
Paleozoic strata are dominated by the cyclic carbonate deposits, while the Mesozoic and Cenozoic strata
consist of siliciclastics. Initial sedimentation during the Cambrian time started over an irregular
Precambrian surface (LeFever et al., 1991). The basin began to subside during the Ordovician Period and
underwent episodic subsidence throughout the rest of the Phanerozoic Period. During the deposition of the
Bakken Formation in the Late Devonian-Early Mississippian, the Williston Basin was an area of active
subsidence (LeFever et al., 1991).

52
Figure 3.1: (a) Map showing the distribution of the Bakken Formation in the Williston Basin and its
equivalent Exshaw Formation and Lower Banff Formation in the Western Canada Sedimentary Basin
(modified from Smith and Bustin, 2000). (b) Locations of the wells used in this study are shown in the
isopach map of the Upper Bakken Shale in the U.S. portion of the Williston Basin (source: Stephen
Sonnenberg, personal communication). Abbreviations of the names of the wells used in this study are listed
in Table 3.1.

53
Figure 3.2: Generalized stratigraphic column of the Williston Basin with an expanded section showing the
stratigraphic details along with the conodont zones of the Late-Devonian-Early Mississippian age Bakken
Formation (Sandberg et al., 1988; Karma, 1991; Hartel et al., 2012; Sonnenberg, 2015).

The paleogeographic reconstruction shows that the Williston Basin was located near the equator
during the deposition of the Bakken Formation and the basin was inundated by a large shallow
epicontinental sea, which had a very low gradient and extended over a large area (Parrish, 1982; Golonka
et al., 1994; Smith and Bustin, 1998; Stasiuk and Fowler, 2004). As shown in Figure 3.2, the Bakken
Formation lies unconformably over the Three Forks Formation and is underlain by the Lodgepole
Formation (LeFever et al., 2011). Bakken Formation is subdivided into three members namely, the organic-
rich Upper and Lower Bakken shales, and the mixed siliciclastic dolomitic siltstones of Middle Bakken
(LeFever et al., 2011). However, recently the “Sanish” member has been included as a part of the Bakken
Formation and has been renamed as the “Pronghorn” Member (LeFever et al., 2011; Johnson, 2013). The
organic-rich Upper and Lower Bakken shales are believed to have been deposited in a moderately deep
water setting (>200 m), while the siltstones and carbonates of the Middle Bakken were deposited in a
shallow marginal- marine setting (LeFever et al., 1991; Smith and Bustin, 1996; Sonnenberg and

54
Pramudito, 2009; Angulo and Buatois, 2012). Change in the relative sea level resulted in the shifting of the
depositional environment from deep water to shallow water setting (Smith et al., 1995; Angulo and Buatois,
2012). During the deposition of the Bakken Formation, the Williston Basin consisted of two depocenters,
one near the center of North Dakota and the other near Elbow sub-basin in Canada. These two depocenters
were separated by the Regina-Melville and Swift Current Platform (Smith and Bustin, 1998). In Alberta,
the Lower and Middle Bakken grades laterally into the Exshaw Formation, while the Upper Bakken Shale
grades into the basal part of the Banff Formation (Richards, 1989; Smith and Bustin, 2000). Bakken
Formation reaches the maximum thickness of about 150ft in North Dakota (LeFever et al., 1991).

UBS is a world-class source-rock with TOC averaging about 11.5 wt.% in the US portion (Webster,
1984;) and 17.6 wt.% in the Canadian portion (Chen et al., 2009) of the Williston Basin. UBS is the most
geographically extensive member and represents the maximum depositional limit for the Bakken Formation
(Cobb, 2013). The thickness of UBS ranges between 4-25ft reaching a maximum in North Dakota near the
Nesson Anticline at basin center as shown in (Figure 3.1b) (Smith and Bustin, 1998). The deposition of
UBS has continued through the Lower Siphonodella Crenulata conodont biozone (Karma, 1991) as shown
in Figure 3.2. Different interpretations for the depositional setting of UBS ranges from the vast swamps to
a deep offshore marine anaerobic environment (MacDonald, 1956; Kents, 1959; McCabe, 1959; Webster,
1984; LeFever et al., 1991; Smith and Bustin, 1996, 1998). Swamps are an unlikely possibility since the
Bakken shales are dominated by sapropelic-amorphous type II kerogen with a minor presence of type III
kerogen (land plants) localized at the basin margin (Webster, 1984; Jin and Sonnenberg, 2012). The widely
accepted depositional model for UBS proposes that it was deposited by suspension sedimentation in a
restricted basin with a depth of more than 200m, where a stratified water column resulted in an anoxic-
euxinic bottom water condition (Smith and Bustin, 1998). However, Egenhoff and Fishman (2013) have
reported the presence of vertical microburrows and sedimentary structures indicating bedload transport
processes in the UBS. Based on these observations, Egenhoff and Fishman (2013) suggested that UBS was
deposited in a dynamic environment with suboxic to dysoxic bottom water conditions.

3.3 Methodology and Background

This section, firstly, details the various UBS cores used in this study. Thereafter, the experimental
methodology for the acquisition of various data is described. This section also provides the details of
externally sourced data used in this study. Finally, the theoretical background for the analyses of this dataset
is outlined.

55
 Core and Dataset

Sub-surface cores of the UBS from a total of eighteen wells, located in North Dakota and Montana
portion of the Williston basin, were used in this study. Table 3.1 lists these wells, including the abbreviated
names used in this study, as well as the full name, API, and location. Well locations from which these cores
were acquired are shown in Figure 3.1. The cores are available at the Wilson M. Laird Core Library of
North Dakota Geological Survey (Grand Forks, ND), USGS Core Research Center (Denver, CO) and Triple
O Slabbing core repository (Denver, CO). For each of these cores, Table 3.1 lists the acquired and external
data, including their sources. XRF elemental data for the UBS section of cores from sixteen wells were
acquired by a hand-held ED-XRF analyzer at 0.5ft intervals. The XRF data were used as proxies for detrital
siliciclastic, carbonate and biogenic sediments, grain-size variations and redox-conditions for
chemostratigraphic analysis. X-ray diffraction (XRD) data were acquired for well LT at Weatherford
Laboratory, Golden, for qualitatively understanding the mineralogy-element association in the UBS. The
externally sourced XRD data for the BF well was also used for this purpose. Thus, XRD data provided the
reference for XRF-based analysis of the compositional differences between the UBS units.

Table 3.1: Names, abbreviations, APIs, and locations of the wells used in this study are listed. The available
dataset used in this study are also listed. Description of abbreviations and symbol used in the table are as
follows, Core Desc.: Core Description, Thin Sec.: Thin-section, ✓: Acquired in this study, ✕: not available,
✓#: Available from external sources.1-USGS, 2-Fishman et al. (2015), 3-(Jin, 2014), 4-(Kocman, 2014),
and 5-NDGS. Sample for well 4-11-7-13W2 was available from Hui Jin (personal communication)
Well Well Full Name API State/ Core XRF Thin TOC XRD
Country Desc. Sec.
LT RR Lonetree Edna 250832269500 MT ✓ ✓ ✓ ✓ ✓
RT Roberts Trust 1-13H 330250098200 ND ✓ ✓ ✓ ✓ ✕
BN BN 1-23H 330070066600 ND ✓ ✕ ✓1 ✓2 ✓2
TT 5-1 Thompson Unit 330070118500 ND ✓ ✕ ✓1 ✓2 ✓2
BF Braaflat 11-11H 330610064100 ND ✓ ✓ ✓ ✓3 ✓4
S Sidonia 1-06H 330610088400 ND ✓ ✓ ✕ ✓3 X
RM Rasmussen 1-21-16H 331050220400 ND ✓ ✓ ✕ ✓5 ✓5
R Rolf 1-20H 331050210000 ND ✓ ✓ ✕ ✕ ✕
C Charlotte 1-22H 330530335800 ND ✓ ✓ ✕ ✕ ✕
L Linseth 13-12 330530369300 ND ✓ ✓ ✕ ✓5 ✓5
T Titan F-WP 32-14-H 330530271100 ND ✓ ✓ ✕ ✕ ✕
RP Round Prairie 1-17H 331050174800 ND ✕ ✓ ✕ ✕ ✕
M Miller 34X-9 330570003500 ND ✕ ✓ ✕ ✕ ✕
WT Wayjetta 44-0311H 330610271700 ND ✕ ✓ ✕ ✕ ✕
J Jorgensen 1-15H 330130149100 ND ✓ ✓ ✕ ✕ ✕
ST Saratoga 12-1 330130166700 ND ✕ ✓ ✕ ✓5 ✓5
G Graham USA 1-15 330070069000 ND ✓ ✓ ✕ ✕ ✕
BY Bullwinkle Yahoo 250832283700 MT ✕ ✓ ✕ ✕ ✕
- Well 4-11-7-13W2 N/A Canada ✕ ✓ ✕ ✓ ✓

56
 Core description was performed at a resolution of 0.5ft for thirteen of the total eighteen cores (Table
3.1). Among these thirteen cores, two wells RT and LT were considered as key wells for this study because
of their strategic locations at the depocenter and basin margin, respectively (Figure 3.1). For the RT and LT
cores, twenty-nine ultra-thin sections were prepared at Weatherford Laboratories, Golden for petrographic
study. Twenty-one ultra-thin sections were perpendicular to bedding, whereas four were both perpendicular
and parallel to bedding. Two additional thin sections were prepared for the immature UBS samples from
the Canadian portion of the Williston Basin. All thin sections were impregnated with blue epoxy and stained
with alizarin red and potassium ferrocyanide to distinguish calcite, dolomite, and ferroan-dolomite. An
additional 23 thin sections from wells TT, BN, and BF were available from the USGS and Bakken
Consortium, CSM. Source rock analysis (SRA) was also performed on the cores of RT and LT wells. High-
frequency SRA for two wells were available from Jin (2014), and SRA for 2-3 samples for four other wells
were available from NDGS.

ED-XRF Instrument Procedure and Interpretation Background

This study used ThermoScientific Niton XL3t GOLDD+ handheld energy dispersive x-ray
fluorescence (ED-XRF) analyzer for acquiring elemental concentration data on the UBS cores. The
analyzer was available at the Colorado School of Mines. The handheld ED-XRF instrument provided a
nondestructive method for semi-quantitative elemental data acquisition. Niton ED-XRF includes three
built-in spectral deconvolution algorithms, or modes: Mining, Soils, and TestAllGeo(TAG)
(www.thermoscientific.com/niton). The Mining and Soils modes are suitable for the measurement of major
and trace elements, whereas the TAG mode is a hybrid of the Mining and Soils modes and is designed for
samples of the largely unknown composition. TAG mode gives the concentration of major, minor, and trace
elements of a given sample in a single measurement. The study required concentrations of these three
categories of elements; therefore, TAG mode was used for all the XRF data acquisition.

Procedures and results for optimum detection time, repeatability, and accuracy analysis for the XRF
measurements are described in Appendix-C. These analyses used thirteen UBS matrix-specific standard
cups, which were prepared following the methodology of Hall et al. (2011). The optimum detection time
for samples used in this study was determined following procedures outlined by Kocman (2014), Lash and
Blood (2014), and Nakamura (2015) and was 180 seconds in the TAG mode. Subsequently, the repeatability
of the XRF measurements was analyzed by the procedure outlined by Fisher et al. (2014) and Nakamura
(2015). The repeatability test indicated that variability in elemental concentration measurements using XRF
was within 1-6%. XRF measurement accuracy for the twelve elements Si, Al, Ti, K, Ca, Fe, Rb, Sr, Mo, U,
V, and Ni in UBS samples was confirmed by comparing the XRF results with the ICP-MS/ICP-OES and
Costech ECS4010 elemental analyzer-derived measurements. For elements Mo, U, V, Ni, Ca, Fe, S, and

57
K, for which the regression coefficient (R2) was more than 0.80, the XRF measurements were considered
applicable for both quantitative and semi-quantitative analyses. Elements Al, Si, Ti and Rb, for which the
regression coefficient (R2) was 0.6-0.80, the XRF measurements were considered applicable for semi-
quantitative analyses only. The log view of the elemental concentration shows similar stratigraphic trends
between the ICP-MS and XRF measurements, which further validates the applicability of XRF
measurements.

In this study, elemental concentrations were acquired on flat core surfaces at 0.5ft intervals for
sixteen wells. The XRF data acquisition on the core samples was performed in the following steps: 1) a
system precheck was run on the XRF analyzer to verify normal operation; 2) the previously described
repeatability procedure (Appendix-C) was repeated after every 25 XRF measurements on the core samples
to confirm whether the R2 for ICP-MS and XRF measurements for the standards were consistent during the
XRF acquisition of the samples, and 3) XRF measurements were taken on flat core surfaces.

Detrital, Biogenic, and Redox Proxies

XRF data were used in this study for chemostratigraphic analysis by utilizing elemental
concentrations as proxies for detrital siliciclastic, carbonate and biogenic sediments, grain-size variations,
and redox conditions. Ti always has a detrital origin, and is present in minerals such as ilmenite, rutile, and
augite in the sand and silt-sized grains; Ti also exists in clays as cations (Calvert, 1976). Elements associated
with siliciclastic sediments are considered to be of detrital origin if they plot with a positive correlation with
Ti. Al is considered to be the principal conservative proxy for clay minerals in fine-grained clastic deposits
(Arthur and Sageman, 1994; Calvert and Pedersen, 2007). K is usually associated with clay minerals,
especially illite or with K-felspar; therefore, after ascertaining its detrital origin by correlating it with Ti, K
was used along with Al as a proxy for siliciclastic detrital sediments. Moreover, a good covariation between
Al, K, and Ti suggested that these elements are primarily associated with illite clay (Wedepohl, 1971).

Zr is preferentially enriched in coarser grain size fraction and is relatively immobile, such that its
concentration does not change due to diagenesis or post-depositional weathering processes. Element Rb is
commonly associated with clay minerals (Turner et al., 2015); therefore, in this study, Zr was used as a
marker for chemostratigraphic correlation purposes, and the Zr/Rb ratio was used as a grain size proxy to
determine variations in the proportion of silt-sized sediments. Si can have both detrital and biogenic and
authigenic origins (Ross and Bustin, 2009). A negative correlation between Si and Ti implies the non-
detrital origin of silica; if this is the case, Si can be used as a proxy for biogenic and/or authigenic sediment.

This study used the nomenclature proposed by Tyson and Pearson (1991) to describe the redox
conditions. According to this nomenclature, oxic conditions refer to 8.0-2.0 ml/1 O2 in water, dysoxic

58
conditions refer to 2.0-0.2 ml/l O2 in water, suboxic conditions refer to 0.2-0.0 ml/1O2 in water, and anoxic
conditions refer to 0.0 ml/I O2 and 0 ml/l H2S in water. Many of the trace metals, such as Mo, U, and V,
exhibit a variation in oxidation state, solubility, and/or change in host-phase as a function of redox condition
of the depositional environment (Tribovillard et al., 2006). These redox-sensitive trace elements are more
soluble and display a conservative behavior in oxic conditions, but are less soluble in reducing conditions,
resulting in their authigenic enrichments in oxygen-depleted facies. This allows them to be used as proxies
for evaluating redox conditions in paleomarine systems (Algeo and Maynard, 2004; Tribovillard et al.,
2006). Euxinic or anoxic-sulfidic water lacks O2 and contains free H2S (Raiswell and Berner, 1985).

According to Tyson and Pearson (1991), the term “bottom water” refers to a water column less
than one meter above the sediment-water interface. In oxic conditions, trace metal Mo in seawater is present
in stable and unreactive molybdate ions (MoO − ). Marine sediments accumulating in oxic environments
have limited authigenic enrichment of Mo, and they exhibit typical detrital background values of Mo
concentration, ranging from 1-5 ppm (Zheng et al., 2000; McLennan, 2001; Morford et al., 2009). In a
reducing environment and in the presence of very small amounts of dissolved HS − , MoO −
ions become
reactive and convert to a series of particle-reactive thiomolybdate ions (MoO� S −x ) (Helz et al., 1996; Zheng
et al., 2000). Mo is rapidly sequestered by sulfurized organic matter, or becomes a solid-solution with Fe-
S and precipitates as authigenic sulfide minerals (Helz et al., 1996; Tribovillard et al., 2004). At higher
levels of dissolved hydrogen sulfide H S a in an euxinic condition, authigenic uptake and burial rate of
Moauth increases by two to three times (Scott and Lyons, 2012); therefore, relative enrichment of Mo usually
indicates the variation in the degree of reducing conditions in the bottom water.

Trace metal U is present mainly as soluble U(VI) in oxic-suboxic conditions, which reduces to less
soluble U(IV) in an oxygen-depleted condition. U concentration in average shale is around 3.7 ppm
(Wedepohl, 1991). Reduction of U commences at the Fe (II)- Fe(III) redox boundary and enrichment of U
is controlled by microbially-mediated Fe redox reactions, rather than by the presence of HS- (Zheng et al.,
2002). The uptake of reduced U� ℎ by the sediment primarily occurs through the formation of organic
metal ligands, or by precipitation of crystalline uraninite or a metastable precursor to uraninite
(Klinkhammer and Palmer, 1991; Zheng et al., 2002). Concentration of trace metal V in average shale is
130 ppm (Wedepohl, 1971, 1991). In oxic waters, V is present as V(V) in the quasi-conservative form of
−
vanadate oxyanions (HVO and H VO− ) (Tribovillard et al., 2006). Under mildly reducing conditions, V(V)
is reduced to V(IV) and forms vanadyl ions (VO − ), hydroxyl species VO OH − , and insoluble hydroxides
VO OH . Under more strongly reducing or euxinic conditions, the presence of free H S , released by
bacterial sulfate reduction, causes V to be further reduced to V(III), which can be taken up by geoporphyrins
or be precipitated as the solid oxide V O or hydroxide V OH phase. The two-step reduction process of V

59
may lead to the formation of separate V carrier phases of contrasting solubilities under non-sulfidic anoxic
versus euxinic conditions (Algeo and Maynard, 2004; Tribovillard et al., 2006). This may result in relatively
high enrichment of total V even when the conditions are non-sulfidic anoxic (Algeo and Maynard, 2004).
High enrichment of V may also occur when anoxic conditions are present only within the pore water while
the bottom water was not necessarily anoxic (Lewan and Maynard, 1982).

I have used variation trends of redox-sensitive trace elements Mo, U, and V, along with other
proxies for identifying the chemostratigraphic packages in the UBS. Redox condition variations in the basin
during the deposition of UBS were inferred from the absolute concentration of Mo. Along with the redox-
sensitive trace elements, the degree of pyritization (DOP) was also used as a proxy to determine the redox
condition of the basin during the deposition of the UBS. DOP is the ratio of pyritic iron (Fe) to reactive iron
(pyritic iron plus HCl soluble iron) and is considered to be a reliable indicator of the depositional redox
conditions in environments which are not limited by reactive Fe (Raiswell et al., 1988; Jones and Manning,
1994). DOP values less than 0.42 indicate normal marine oxygenated conditions, values ranging from 0.42
to 0.75 indicate dysoxic-anoxic conditions with a less strongly stratified water column, and values greater
than 0.75 indicate anoxic conditions with a strongly stratified water column (Raiswell et al., 1988; Hatch
and Leventhal, 1992; Algeo and Maynard, 2004). Uniformly high DOP values typically indicate deposition
under euxinic conditions and suggest that nearly all of the available reactive Fe was utilized in the formation
of pyrite (Raiswell et al., 1988; Lyons and Berner, 1992). Although the calculation of DOP values requires
the determination of acid-soluble Fe and sulfide S concentrations via the chromium reduction method,
Algeo and Maynard (2004) suggested that total concentration of Fe and S can be used to make an
approximate assessment of total DOP or DOPT values from the crossplot of Fe (total) and S (total). Algeo
and Maynard (2004) used DOPT to infer the redox condition of Pennsylvanian Kansas cyclothems. I have
determined threshold DOPT values for dysoxic, anoxic, and euxinic conditions by following the technique
of Algeo and Maynard (2004) through the crossplot between Fe (total) and S (total), and used them as redox
proxies.

Core Description and Petrography

Core description and petrographic analysis were performed in the mudstone-dominated succession
of the UBS to determine the lithofacies variability in both hand-specimen (mm-cm) scale and microscopic
(sub-mm) scale. All thin sections used in this study were scanned in a flat-bed scanner (HP CLJM477) at a
resolution of 300 dpi to capture 1-10 mm-scale textures. Thereafter, the thin sections were studied using
optical microscopy to obtain 0.1-1 mm-scale features. Micron-scale textural details were obtained for four
of the thin sections by applying carbon coating and then examining them under a scanning electron
microscope (SEM) using back-scattered electron (BSE) detector and energy dispersive spectroscopy

60
(EDAXTM). SEM analysis was performed using the TESCAN MIRA3 LMH Schottky Field-Emission SEM
at the Colorado School of Mines.

The lithofacies present in the UBS were described based on their compositional and textural
features including grain size, grain origin (allochthonous, autochthonous, or diagenetic), sedimentary
structure, and bedding geometries. I used the nomenclature scheme of Macquaker and Adams (2003) as
follows: mudstones containing more than 90% of a particular component are termed as ‘dominated’ by that
particular constituent; those containing 50-90% of a particular component are termed as ‘rich’ in that
constituent; and those containing 10-50% of a constituent are termed as ‘bearing’ that constituent. The
nomenclature of each facies of the UBS was finalized by combining the observations of core descriptions
and microscopic petrography studies.

SRA and ICPMS Measurements

TOC was measured on all of the fifty core chips from both the RT and LT cores using the
Weatherford Source Rock Analyzer (SRA) instrument at Colorado School of Mines, using the procedure
described by Jin (2014). Prior to SRA, the samples were cleaned with deionized water and air-dried
overnight to remove salt crust and drilling mud residues. Subsequently, the cleaned sample pieces were
crushed by mortar and pestle to 200-mesh size powder. Aliquots of the powdered samples were then
analyzed for TOC. Remaining aliquots of the powdered samples were used for determining the
concentration of the minor and trace elements using inductively coupled plasma mass spectrometry (ICP-
MS) at SGS Services, Canada. The ICP-MS results were used for making matrix-specific UBS standards
for validation of ED-XRF-derived results. ICP-MS results will also be used in Chapter-4 for understanding
the relationship of TOC to different geochemical proxies.

3.4 Results and Interpretation

In this section, I identify various lithofacies of the UBS and interpret their depositional processes.
I then present the results of the chemostratigraphic analysis and describe the various chemostratigraphic
units of the UBS and their regional correlations. Finally, I describe the distribution of various facies in the
chemostratigraphic units.

Facies Description and Depositional Processes

Based on both compositional and textural features, six lithofacies of the UBS were identified. The
UBS comprises of a variety of intergradational lithofacies, which are dominated by fine-grained (clay to
silt-size) sediments, except for phosphatic skeletal lag beds. The six identified lithofacies of UBS are as
follows: (i) siliceous mudstone (F1); (ii) massive to finely laminated silt-bearing mudstone (F2); (iii) sand

61
and clay-bearing silt-rich mudstone (F3); (iv) macrofossil-bearing silt-rich mudstone (F4); (v) burrow-
mottled silt-bearing mudstone (F5); and (vi) phosphate and fossil-rich mudstone (F6). Out of these six
facies, 95% of the UBS succession consisted of the three facies F1, F2, and F4. Facies F1 was the most
dominant facies in the UBS, followed by facies F2. The presence of facies F4 in the UBS is localized. The
remaining three facies make up less than 5% of the entire UBS. The next sub-sections describe
characteristics of each of the facies.

Facies F1- Siliceous mudstone

Facies F1, which is characterized by the abundance of radiolarian tests, occurs as 1-40 mm thick
mudstone beds intercalated within massive to finely-laminated silt-bearing mudstone (F2). This facies was
further divided into two sub-facies: (i) Radiolarites, or massive radiolarian-dominated siliceous mudstone
(F1a), as shown in Figure 3.3a, and (ii) Finely laminated radiolarian and silt-bearing mudstone (F1b) as
shown in Figure 3.3b. Facies F1a are characterized by their massive fabric and are composed primarily (up
to 90%) of radiolarian tests and spines (Figure 3.4a and b), which in turn are filled with microcrystalline
silica, pyrite, sphalerite and/or calcite (Figure 3.4c and d). The matrix is composed of authigenic
microcrystalline silica, pyrite cement, and organic matter, which gives the appearance of diffused wispy
aggregates against the radiolarian tests. In some radiolarite beds, only a diffused outline of the radiolarian
body exists, while the radiolarian tests appear to have dissolved out (Figure 3.4d).

Figure 3.3: (a) Core photograph illustrating massive radiolarian-dominated siliceous mudstone (F1a). The
characteristic ptygmatic pyrite-filled, sub-vertical fractures in facies F1a are shown by white arrows. Well:
Rasmussen 1-21-6H, 10211ft. (b) Core photograph illustrating finely laminated radiolarian and silt-bearing
mudstone (F1b). Well: Rasmussen 1-21-6H, 10214ft.

62
Figure 3.4: (a) Scanned petrographic thin section image showing ~14mm thick bed of massive radiolarian
dominated mudstone (F1a). The sharp contact at the base and top of the facies F1a are indicated by the
white arrow; Well: LTE, 10386 ft. (b) Optical photomicrograph of F1a showing the typical fabric of the
massive radiolarian dominated siliceous mudstone (F1a). The high content of radiolarians, which are shown
by the red arrow is evident from this photomicrograph. Wispy organic matter surrounding the radiolarians
is shown by the yellow arrow. (c) Thin section photomicrograph of facies F1a showing radiolarian tests,
which was replaced by chert (green arrow), sphalerite (yellow arrow) and pyrite (red arrow). Well: RT,
10663.60 ft. (d) Optical photomicrograph of facies F1a showing is radiolarians, which are replaced by
calcite (green arrow) and pyrite (red arrow). As evident, it is difficult to distinguish the shape of the
individual radiolarians as they are mostly dissolved out. Well: RT, 10661.9 ft.

Facies F1a generally has a sharp basal and top contact. The thickness of facies F1a occasionally
varies due to diagenetic overprint and is related to the silica dissolution and formation of early authigenic
pyrite (Figure 3.3a) (Vecsei et al., 1989). The characteristic sub-vertical calcite, pyrite and/or sphalerite-
filled ptygmatic fractures along with bitumen-filled horizontal fractures are frequently observed features in
facies F1a and represent fluid-escape structures (Vecsei et al., 1989)(Figure 3.5a and b). Moreover, this
facies was found to have the lowest TOC, ranging from 1-6 wt.%. Facies F1b is characterized by varying
amounts (30-50%) of radiolarians, which are either dispersed randomly along with silt grains in an organic
matter-bearing and clay-rich matrix or forms single test thick continuous to discontinuous laminae (Figure

63
3.6a and b). Similar to F1a, the characteristic sub-vertical calcite, pyrite and/or sphalerite-filled ptygmatic
fractures are also present in F1b, but their occurrence is comparatively less frequent. The upper and lower
contacts of facies F1b are commonly gradational. The TOC of facies F1b is more than those of facies F1a,
F4, and F5, but comparatively lower than that of facies F2.

Figure 3.5: (a) Optical photomicrograph showing a vertical ptygmatic fracture (red arrow) in massive
radiolarian dominated mudstone (F1a) Well: RT, 10663.60ft. (b) Optical photomicrograph showing the
vertical fracture (red arrow) in Figure 3.5a at higher magnification. Two generations of cement within the
fracture, sphalerite (red arrow) followed by pyrite (yellow arrow) is observed. Well: RT, 10663.6ft.

The presence of radiolarians indicates sediment accumulation in the outer shelf to deep basinal
offshore settings at a water depth of more than 50m (Stasiuk and Fowler, 2004). Therefore, the occurrence
of facies F1 indicates that the UBS was deposited in a setting, which had more than 50m of water depth.
Facies F1a is interpreted to represent episodic radiolarian blooms related to eutrophication pulses, such that
the radiolarians were rapidly deposited by suspension sedimentation (Garrison and Fischer, 1969; Jones
and Murchey, 1986; De Wever and Baudin, 1996; De Wever et al., 2002). This interpretation is supported
by the near absence of detrital sediments, massive structures, and the presence of sharp basal and top
contact. The deposition of facies F1b represents the background sedimentation of organic matter,
radiolarians, and fecal pellets along with detrital clay and silt which continued after episodes of radiolarian
blooms. The single-test thick laminae of radiolarian tests in facies F1b is interpreted to represent shorter
periods of radiolarian blooms during a constant rate of sedimentation of detrital sediments and organic
matter (Jones and Murchey, 1986). Such radiolarian blooms may be quasiperiodic and related to seasonal
fluctuations in nutrient supply.

Alternatively, if the single-test thick radiolarian laminae were deposited due to the hydrodynamic
sorting of radiolarian tests by bottom water currents, it would have been a dynamic bottom water

64
environment (Vecsei et al., 1989). This would result in oxic to suboxic conditions at the sediment-water
interface; however, lack of any bioturbation suggests that benthic conditions were not conducive for the
survival of benthic organisms; therefore, it represents highly reducing conditions. This suggests that the
benthic condition was stagnant, and bottom currents were sluggish and occasional (Smith and Bustin, 1996).
The absence of any other bioclasts, coarse-grained sediments, or bedload transport sedimentary structures,
such as ripples or grading, further indicates that the single-test thick radiolarian laminae were deposited by
suspension sedimentation and were not produced by bottom current activities (Jones and Murchey, 1986;
Vecsei et al., 1989). Therefore, deposition by predominantly suspension sedimentation of planktonic
organisms, and lack of any major sedimentary structures indicating bedload transport processes and
bioturbation suggest that benthic conditions were most-likely anoxic-euxinic during the deposition of facies
F1.

Figure 3.6: (a) Scanned petrographic thin section image of laminated radiolarian and silt-bearing mudstone
(F-1b). The double-headed yellow arrows show sub-mm scale laminae of F1b. Red arrows showing
sphalerite-filled vertical fractures in the middle of the image Well: LT, 10390.1ft. (b) An optical
photomicrograph of laminated radiolarian and silt-bearing mudstone (F-1b) showing discontinuous laminae
of radiolarians (red arrows). Radiolarians (yellow arrows) along with silt grains are observed to be randomly
distributed in an organic matter and clay-rich matrix in the lower part of the image. Well: LT, 10390.1ft.

Facies F2- Massive to finely laminated silt-bearing mudstone

Facies F2 is massive to finely laminated in character, with gradational contacts, and is characterized
macroscopically by the presence of mm-thick discontinuous pyrite lenses which occur abundantly along
the fine laminations (Figure 3.7a). This facies is also characterized by the presence of detrital silt grains,
fecal pellets, and organomineralic aggregates (a term defined by Macquaker et al., 2010; Ghadeer and
Macquaker, 2012) as the major textural components (Figure 3.7b). Facies F2 was found to be the most

65
abundant facies type in the UBS and makes up about 75% of the entire UBS. The thickness of this facies
can be up to 4ft.

Figure 3.7: (a) Core photograph showing the massive to finely laminated silt-bearing mudstone (F2).
Discontinuous mm-size pyrite lenses (shown by the yellow arrow) is the most distinct macro-scale
characteristics of F2. Well: R, 1265ft. (b) Optical photomicrograph of massive to finely laminated silt-
bearing mudstone (F2) showing the presence and fecal pellets (p) and silt grains which are randomly
dispersed in the matrix. Well LT, 10384.6ft, TOC: 14.3 wt.%. (c) Optical photomicrograph showing
massive to finely laminated silt-bearing mudstone (F2a), which have characteristically high matrix content
and proportionately less fecal pellets (<5%) compared to massive to finely laminated fecal pellet and silt-
bearing mudstone (F2b). Well: BN, 10352.1ft. (d) Optical photomicrograph showing massive to finely
laminated fecal pellet and silt-bearing mudstone (F2b), which have proportionately higher amounts of fecal
pellets (p) (>5%) compared to massive to finely laminated silt-bearing mudstone (F2a). Well: 4-11-7-13W2,
1776.25ft, TOC: 28.6 wt.%.

Based on the proportion of fecal pellets and the amount of matrix, I divided this facies further into
two types: silt-bearing mudstone (F2a) (Figure 3.7c) and fecal-pellet and silt-bearing mudstone (F2b)
(Figure 3.7d); however, facies F2a and F2b are distinguishable only under a microscope. Silt content in

66
both the sub-facies varies between 10-25%; however, in facies F2a, the matrix, which is composed of
organic matter and clay minerals, varies between 65-85%, and fecal pellets are less than 2%. In facies F2b,
the organic matter-bearing and clay-rich matrix may vary between 35-85%, and fecal pellets may vary from
5-45%. In both sub-facies F2a and F2b, silt is primarily composed of angular to sub-angular quartz, feldspar
grains, and detrital and authigenic dolomite, along with pyrite framboids (Figure 3.8a). Silt grains are
randomly dispersed in the clay and organic matter-rich matrix. Occasionally, radiolarians, conodonts, and
phosphatic bones are also present, making up less than 5% of all framework components (Figure 3.8b).

Facies F2 was interpreted to be deposited by suspension settling of organominerallic aggregates,

fecal pellets, clay minerals, and silt grains on the basis of their fine planar laminations and the absence of
sedimentary structures, which, in turn, indicates the absence of bedload transport processes. High TOC,
along the with the absence of bedload transport features and a lack of burrowing, indicated that the bottom-
water conditions were predominantly anoxic to euxinic during the deposition of this facies. As shown in
Figure 3.7b and d, the uniform size, appearance, and abundance of type 1 fecal pellets, which are OM-poor
and clay-rich, suggested that they were most likely produced by a single type of organism (Macquaker et
al., 2010; Lohr and Kennedy, 2015). Benthic macrofauna cannot thrive in anoxic conditions; therefore, the
type 1 fecal pellets, which dominate this facies, were either produced by filter-feeding zooplanktons living
in the oxic water column, or by deposit-feeding benthic meiofauna, which can survive in anoxic to euxinic
conditions at the sediment-water interface (Cuomo and Bartholomew, 1991; Neira et al., 2001; Turner,
2002; Macquaker et al., 2010; Lohr and Kennedy, 2015).

Figure 3.8: (a) FE-SEM image of massive to finely laminated silt-bearing mudstone (F2) showing the
presence of pyrite framboids (pyr fram), which are less than 10µm in size and are randomly dispersed in
clay and organic matter-rich matrix. Well: RT, 10668ft, TOC:14.7 wt.%. (b) Optical photomicrograph
showing the occasional presence of radiolarians (rad) in massive to finely laminated silt-bearing mudstone
(F2). Well: RT, 10665.5ft, TOC: 12 wt.%.

67
 As shown in Figure 3.9a and b, two types of fecal pellets were identified in facies F2. The dominant
type 1 fecal pellets are poor in organic matter and rich in clay and have a distinct homogeneous texture
compared to the matrix (Figure 3.9a, c, and d). They are cylindrical in shape with tapered ends and 100-
500 µm in length. The rare type 2 fecal pellets are silt-rich, circular-shaped, and are generally up to 350 µm
in length, as shown in Figure 3.7b. The TOC in the immature samples of this facies from the Canadian
portion of the UBS was as high as 28.6 wt.% and was the highest TOC among all the facies studied.

Figure 3.9: (a) Optical photomicrograph from a parallel to bedding thin-section from massive to finely
laminated silt-bearing mudstone (F2) showing the cross-section of type 1 fecal pellet, which is cylindrical
in shape, clay-rich and organic poor. Well: LT, 10388.1ft, TOC: 11.2 wt.%. (b) Optical photomicrograph
from a parallel to bedding thin-section from massive to finely laminated silt-bearing mudstone (F2) showing
the cross-section of type 2 fecal pellet, which is circular-shaped and silt-rich. Well: LT, 10388.1ft, TOC:
11.2 wt.%. (c) FESEM image showing the shape of type 1 fecal pellet, which is cylindrical in shape and
200µm in length. Well: RT, 10665.6ft, TOC: 10.5 wt.%. (d) The SEM-EDAX image showing the near
absence of organic matter (red) within the type 1 fecal pellet, while the matrix is rich in organic matter.
Well: RT, 10665.6ft, TOC:10.5 wt.%.

68
 The type 1 fecal pellets were identified as planktonic in origin based on the following observations
– (i) cylindrical shape, and the tapered end of the type 1 pellets are characteristically produced by planktonic
organisms (Cuomo and Bartholomew, 1991; Lohr and Kennedy, 2015), and (ii) continuous wispy organic
matter and organominerallic aggregates along the fecal pellets indicate that the organic matter has not been
disrupted by meiofaunal burrowing (Lohr and Kennedy, 2015). Therefore, the presence of planktonic fecal
pellets indicates that the surface water was well oxygenated, but the bottom water condition was not
conducive for living organisms. The presence of abundant pyrite framboids further supports the hypothesis
that the bottom water condition was anoxic to euxinic during the deposition of facies F2.

Facies F3- Sand and clay-bearing, silt-rich mudstone

Facies F3 is characterized by coarser framework components, consisting of very fine sand and silt-
sized detrital grains of quartz, dolomite, and feldspar that are organized into sub-mm to 3 mm-thick laminae
with a sharp base and a commonly gradational top, shown in Figure 3.10a, 10b, and 10c. The amount of
detrital silt and sand varies from 30-60% in this facies. Phosphatic clasts are also present along with the
detrital grains; however, their volume is less than 5% (Figure 3.10d) The volume of the matrix, which is
composed of organic matter and clay minerals, varies considerably in the clay-rich laminae and in the silt-
rich laminae. Facies F3 exhibits continuous planar laminations, starved ripples (Figure 3.10e), the lenticular
fabric of coarse silt laminae (Figure 3.11a) and normal grading (Figure 3.10d). The normally graded beds
with sharp bases are mostly overlain by silt-rich laminae, which grades upwards into more clay-rich
laminae. Moreover, occasional horizontal burrows are also present in the clay-rich laminae. Facies F3
grades into massive to finely laminated silt-bearing mudstone (F2). As these mudstones are thin-bedded,
accurate determination of TOC is difficult without contaminating the sample with facies F2.

Facies F3 is interpreted to be deposited by bedload transport processes. The lenticular nature of the
coarse silt laminae and starved ripples indicate that these mudstones were likely deposited by storm-induced
bottom water currents which carried silt grains and flocculated mud in bedload (Schieber et al., 2007;
Schieber, 2011, 2016). Occasionally, phosphatic fragments accumulated along silty laminae, which
indicates a weak or short duration of bottom current events (Li and Schieber, 2015). The presence of
phosphatic clasts, along with the detrital silt and sand as normally graded beds with the micro-scoured base,
suggest that some of these mudstone beds were probably deposited as distal tempestites by waning and
storm-induced combined flows (Aigner and Reineck, 1982; Ghadeer and Macquaker, 2011; Schieber,
2016). The occurrence of rare horizontal burrows Planolites in the clay-rich laminae in Facies F3 indicates
that after the deposition of these beds, the sediment-water interface was suboxic for a limited time. Because
of this, the clay-rich tops of these units were colonized by diminutive organisms.

69
Figure 3.10: (a) Core photograph showing continuous and discontinuous lamination (yellow arrow) of sand
and clay-bearing, silt-rich mudstone (F3). (b) Scanned image of a petrographic thin section showing planar
silt laminations (yellow arrow) in the sand and clay-bearing, silt-rich mudstone (F3) well-LT, depth-10389ft
(c) Optical photomicrograph showing framework components of very-fine sand to coarse silt grains in a
clay and organic-rich matrix in facies F3. (d) An optical micrograph showing massive to finely graded beds
of facies F3. Presence of phosphatic fragments is indicated by the yellow arrows. Well: LT, 10387.1ft. (e)
The optical photomicrograph showing starved ripples and continuous planar silt lamina. Well: LT,
10389.8ft.

70
Figure 3.11: (a) Optical photomicrograph showing lenticular silt beds with a sharp base and overlain by a
clay-rich lamina. The sample is from well LT, depth 10389ft. (b) The optical photomicrograph is showing
horizontal Planolites burrow indicated by yellow arrow in the clay-rich lamina overlying the silt lamina.
Well: RT, 10670ft.

Facies F4- Macrofossil bearing silt-rich mudstone

Facies F4 is massive in character and is macroscopically distinguishable due to the presence of
horizontally aligned, articulated and disarticulated brachiopod shells up to 10 mm in length, making up to
10% of the total volume of this facies, as shown in Figure 3.12a-d. Facies F4 has a mostly homogeneous
texture and has the highest detrital silt content, varying from 20-60%, shown in Figure 3.13a and b. Detrital
silt grains are primarily composed of quartz, dolomite, and felspar. Additionally, this facies is characterized
by the presence of agglutinated benthic forams, demonstrated in Figure 3.13a and b. As indicated in the
Figure 3.13c and d, the benthic forams observed in this study were either cherty or had silt-size quartz
grains in their walls. Detrital silt grains, shell fragments, and benthic forams are randomly distributed in the
F4 matrix, which is composed of clay minerals and organic matter. Figure 3.12c and d show that calcareous
shells present in this facies were often replaced by pyrite and silica. Figure 3.14a shows that occasionally,
a few silt grains-thick laminae exhibiting discontinuous planar or undulating geometries with downlapping
surface were also observed in facies F4. Moreover, as shown in Figure 3.14b, macro-scale Planolites
burrows, which are 2-5mm in size, were occasionally present in this facies. The maximum thickness of this
facies is as high as 6ft. The average TOC of facies F4 varies from about 3-8 wt.%, which is comparatively
lower than that in facies F2.

71
Figure 3.12: (a) Core photograph of macrofossil-bearing silt-rich mudstone (F4) showing horizontally
aligned thin-shelled brachiopods (brac), which is a diagnostic characteristic of this facies. Well: RT,
10654ft. (b) Core photograph of macrofossil-bearing silt-rich mudstone (F4) showing thin-shelled
brachiopods in facies F4. Well: L, 10937.2ft. (c) The optical photomicrograph showing articulated
brachiopods (brac) in facies F4. Brachiopods are partial to completely pyritized. Well: RT, 10654.5ft. (d)
The optical photomicrograph showing articulated brachiopods (brac) in facies F4. Brachiopods are replaced
by silica and pyrite. Well: RT, 10653.5ft.

Loucks and Ruppel (2007) and Schieber (2009) reported the presence of agglutinated benthic
forams from other Late Devonian - Early Mississippian black shales, such as the Mississippian Barnett
Shale, and the Late Devonian Chattanooga Shale of Tennessee; however, for the UBS, which belongs to
the same age, the presence of benthic forams have not been reported previously, and the likely reason for
this is that facies F4 itself was not identified previously. This study found that benthic forams are abundantly
present as this facies was closely examined through petrographic thin-section study. As the existence of

72
benthic forams requires at least some oxygen to persist at the seafloor (Schieber, 2009), facies F4 was
interpreted to be deposited in a suboxic condition, and the sediment-water interface was occasionally
hospitable, which probably supported the survival of meiofaunal organisms. This is further supported by
the presence of Planolites macroburrows; therefore, the bottom water conditions of facies F4 were probably
suboxic and much less reducing compared to facies F1 and F2. As the benthic conditions were suboxic and
macroburrows are occasionally present in this facies, the homogeneous texture is probably a result of
meiofaunal burrowing (Lohr and Kennedy, 2015). Meiofaunal burrowing likely destroyed the primary
laminations, and thus the depositional process could not be identified.

Figure 3.13: (a) and (b) Optical photomicrograph in plane polars and crossed polars showing the presence
of abundant silt content and homogeneous fabric in facies F4. Homogeneous fabric in facies F4 is probably
a result of meiofaunal burrowing. Abundant detrital dolomite (dd), authigenic dolomite and agglutinated
benthic foram(af) in this facies are observed at the lower right corner of the image. Well: RT, 10653.5ft
TOC: 2.78 wt.%. (c) and (d) Optical photomicrograph in plane polar and crossed polars, showing
agglutinated benthic forams (af) in the facies F4 at a higher magnification. Presence of quartz silt in the
wall of the benthic foram can be seen in the wall of benthic foram as pointed by the yellow arrow. Well-
RT, 10650ft, TOC: 7.34 wt.%.

73
 The infrequent presence of undulating silt laminae with downlapping surface indicated occasional
deposition by bedload transport processes under the influence of weak bottom water currents. Most
articulate-disarticulated brachiopods present in this facies are assumed to have lived in well-oxygenated
surface water conditions, which later settled down to the bottom sediment along with algae or other floating
plant materials (personal communication with Mark Longman). Alternatively, as the benthic condition was
suboxic, some of the shell fragments may belong to bivalve Bositra, which can also survive in soft muddy
benthic condition (Leonowicz, 2016).

Figure 3.14: (a) Optical photomicrograph is showing few silt grains thick discontinuous planar laminae
with downlapping surface (shown by the yellow arrow) in macrofossil-bearing silt-rich mudstone (F4).
Well: RT, 10354.5ft (b) Core photograph showing the presence of macroscale Planolites burrows (pl) in
macrofossil-bearing silt-rich mudstone (F4). Well: L, 10934ft.

Facies F-5- Burrow-mottled silt-bearing mudstone

Facies F5 is characterized by the presence of macro-sized horizontal burrows and is present
primarily at the contact of the UBS with the overlying Lodgepole Formation, (Figure 3.15a and b). As
shown in Figure 3.15, the elliptical-shaped horizontal burrows, which were identified as Planolites, are 2-
10 mm in length and about 0.5-1mm in thickness. This facies consists of around 20-30% detrital silt grains
of quartz, dolomite, and feldspar, dispersed randomly in an organic matter and clay-rich matrix. These
burrows are distinguishable in the darker matrix because of their lighter color, as they are devoid of organic
matter, shown in Figure 3.15d; however, the other constituents (non-organic) of the burrow-fill were similar
to that of the surrounding matrix. TOC of this facies varies considerably from 2 wt.% in core RT to 11 wt.%
in core LT and depends on the amount of bioturbation. Due to bioturbation, the depositional process of this
facies cannot be identified; however, the presence of macro-scale burrows indicates that this facies was
deposited in an oxic to suboxic condition.

74
Figure 3.15: (a) Core photograph showing macro-scale Planolites burrows (b) in burrow-mottled silt-
bearing mudstone (F5) at the contact of UBS with the overlying Lodgepole Formation. Well: LT, 10383.2ft.
(b) A core photograph is showing macro-scale Planolites burrows (b) in burrow-mottled silt-bearing
mudstone (F5) at the contact of UBS with the overlying Lodgepole Formation. Well: RT, 10648ft. (c)
Scanned image of petrographic thin section showing Planolites burrows (b) in facies F5. Well: LT,
10383.2ft. (d) Optical photomicrograph of facies F5 is showing that the burrow (b) filled by constituents,
which are similar to the matrix, and is devoid of organic matter. Well: LT, 10383.2ft.

F-6: Phosphate clast and fossil-rich mudstone

Facies F6 is massive in character and is distinguished by an abundance of phosphatic fragments,
conodonts, fish bones, and sponge spicules, which collectively form 20-70% of the total volume of this
facies (Figure 3.16a-d). This facies occurs as 0.5-10 mm thick beds with mostly gradational bases. Basal
microscouring is occasionally present in this facies, mostly at the contact of the UBS with the underlying
Middle Bakken member (Figure 3.16b). The phosphatic fragments and fossils present in this facies are
randomly oriented and do not show any grading. Detrital siliciclastic and dolomite silt grains form 10-20%
of the volume of the framework of facies F6. The matrix content of this facies varies widely, as the
framework components form a packstone-type texture with no matrix, or it may form a floating texture with

75
an abundant matrix composed of organic matter and clay minerals. This facies is interpreted to be deposited
by the reworking of conodonts, fish bones, sponge spicules, and phosphate grains, which were concentrated
by storm-induced bottom water currents and; therefore, represent lag deposits or distal tempestites (Aigner
and Reineck, 1982; Ghadeer and Macquaker, 2011). The amount of matrix varies in this facies is probably
dependent on the current intensity. Moreover, the absence of significant scouring surfaces suggests that the
storm-induced currents were episodic and low in intensity. Smith and Bustin (1996) observed similar lag
deposits in the Bakken Shale in the Canadian portion of the Williston Basin and concluded that they were
produced by episodic bottom water circulation.

Figure 3.16: (a) Core photograph showing mm-thick bed of phosphate clast and fossil-rich mudstone (F6).
Well: R, 11026.6ft. (b) Core photograph showing cm-thick bed of phosphate clast and fossil-rich mudstone
(F6). Well: BN, 10358.4ft. (c) Core photograph is showing cm-thick bed of phosphate clast and fossil-rich
mudstone (F6) at the contact of UBS with the Middle Baken. Well: R, 11044.6ft. (d) Optical
photomicrograph showing phosphate clasts, and sponge spicules (shown by red arrow) oriented randomly
in phosphate clast and fossil-rich mudstone (F6). The base of this facies shows microscouring. Well: BN,
10352ft.

76
 Factor Analysis and Element Relationships using XRF data

Principal Component Analysis (PCA), which is a type of statistical factor analysis, was performed
on the XRF elemental data of the sixteen wells (Table 3.1). XLSTAT, which is an add-in module of MS-
Excel, was used to perform the PCA for twenty elements: silicon (Si), aluminum (Al), titanium (Ti),
potassium (K), magnesium (Mg), zirconium (Zr), rubidium (Rb), niobium (Nb), iron (Fe), calcium (Ca),
manganese (Mn), strontium (Sr), molybdenum (Mo), uranium (U), vanadium (V), nickel (Ni), zinc (Zn),
copper (Cu), sulfur (S), and chromium (Cr). The crossplot between the two principal components, or the
highest factors, shows a dominant relationship between the twenty elements. PCA results for XRF
elemental data from all sixteen cores show four groups of elements, as shown in Figure 3.17. The four
groups, or clusters, which were identified in the PCA, are as follows: 1) Al, K, Ti, Rb, and Nb; 2) Ca, Sr,
Mn, and Mg; 3) Mo, U, Zn, Cu, Cr, Fe and S; and 4) Si.

Figure 3.17: Four groups of element clusters shown on the crossplot between the two principal components
determined from Principal Component Analysis (PCA) in XLSTAT TM of twenty elements. The PCA was
performed on around 400 UBS sampled from sixteen cores.

From the first group, it was inferred that elements Al, K, Rb, and Nb were detrital in origin since
they cluster with element Ti, which is always associated with detrital minerals (Section 3.3.3). These detrital
elements are commonly associated with detrital siliciclastic minerals such as clays, quartz, and feldspar.
From the second group, it was inferred that elements Ca, Mg, Mn, and Sr are associated with carbonate
minerals. Figure 3.17 shows that Zr plots near the carbonate group of elements. Zr is usually associated

77
with detrital silt or sand-sized sediments, suggesting that some of the carbonate silt may have a detrital
origin in the UBS. From the third group, elements Mo, U, V, Ni, Zn, Cr, and Cu are assumed to be redox
sensitive, and become enriched under reducing conditions (Tribovillard et al., 2006). Elements Fe and S
plot in close proximity within the redox group because of their association with pyrite. V, Cr, and Cu cluster
together and plot closer to the detrital group of elements in comparison to other redox-sensitive elements,
such as U and Mo. Tribovillard et al. (2006) reported that Cr usually has a strong detrital influence;
therefore, this suggests that along with Cr, V and Cu also have some detrital influence in the UBS. Finally,
the fourth group consists of just Si, suggesting that a fraction of Si is of biogenic and/or authigenic origin
as it plots opposite to the detrital group. This further substantiates the presence of the abundant radiolarians
in facies F1 of the UBS as previously discussed in Section 3.4.1. Moreover, as the detrital group of elements
in the UBS mostly reflects clay-sized sediments and Si has a biogenic origin, Zr does not plot with either
of these two groups.

Crossplots between Elements Concentrations

In PCA, Al, K and Ti plot together and form the Group 1 cluster (Figure 3.17), which indicates that
Al and K have a detrital origin. This is substantiated by the strong positive correlation observed between Ti
vs. Al and Ti vs. K crossplots, as shown in Figure 3.18a, b. The Al vs. K crossplot also shows a strong
positive correlation (Figure 3.19a); therefore, both Al and K were used as proxies for detrital clays since Al
is mostly associated with clay minerals as described in Section 3.3.3. PCA suggested that Si in the UBS is
of non-detrital origin, which is verified by a weak negative correlation between Si and Ti in Figure 3.19b.
The scatter in the crossplot in Figure 3.19b is due to both the detrital and biogenic origins of Si in the UBS.
This is consistent with the findings of the petrographic studies, which confirmed the presence of radiolarians
in facies F1 (Figure 3.4a) and detrital quartz silt in facies other than F1 (Figure 3.7). Therefore, a
combination of high Si and low Ti was used as indicators for radiolarian-rich intervals in the UBS.

The UBS was found to be enriched in the redox-sensitive trace elements of Mo, U, and V, as their
values are higher than that in the average shale, as indicated by the box and whisker plot in Figure 3.20.
This suggests UBS has a high enrichment of these redox sensitive trace elements; however, there are
significant variations in the degrees of enrichment (Figure 3.20). Nonetheless, Mo and U show strong
covariance (Figure 3.21a), which suggest that their enrichment occurs in similar proportions in the
prevailing redox conditions. As shown in Figure 3.21b, the positive correlation between Fe and S indicates
that these elements are mainly associated with pyrite. Moreover, the presence of abundant pyrite framboids
was also observed in the petrographic thin sections (Figure 3.8) which further substantiates the association
of these elements with pyrite. Also, both the elements plotted in the redox group 3 in the PCA plot (Figure
3.17), which further validates that Fe and S are associated with pyrite that formed under reducing

78
conditions. Therefore, DOPT, determined from the Fe vs. S plot can be used a proxy to understand the redox
condition of the basin during the deposition of UBS.

Figure 3.18: (a) Crossplot of Al vs. Ti showing a strong positive covariance between the two elements,
which indicates that Al has a detrital origin in UBS. (b) Crossplot of K vs. Ti showing strong positive
covariance between the two elements indicating K has a detrital origin. Each crossplot includes all the
samples from the sixteen wells for, which XRF data were collected in this study. All the data points are
color-coded by the three different chemostratigraphic units of UBS which are described in Section 3.4.5.
All the three crossplots form two clusters, which correspond to the two main chemostratigraphic units,
namely Unit-1 and Unit-2.

Figure 3.19: (a) Cross-plot of Al vs. K showing a strong covariance between the two elements, which
indicates that most of the Al and K are associated with clay minerals in UBS. (b) Crossplot between Ti vs.
Si showing a negative correlation, which indicates that f Si has both non-detrital origin biogenic/authigenic
origin and detrital origin. All the data points are color-coded by the three different chemostratigraphic units
of UBS which are described in Section 3.4.5. All the three crossplots form two clusters, which correspond
to the two main chemostratigraphic units, namely Unit-1 and Unit-2.

The weak covariance between Mo or U with V (Figure 3.22a and b) suggests that enrichment of V
during the deposition of the UBS was dependent on other factors or conditions. This is explained by the
two-step reduction process of V, as previously described in Section 3.3.3. In the first step of V reduction,

79
which takes place in anoxic non-sulfidic conditions, the rate of V enrichment is rapid; however, the anoxic
non-sulfidic conditions are not as conducive for the enrichment of Mo and U, which require sulfidic
conditions for the high degree of enrichment similar to that observed in the UBS. Therefore, in anoxic non-
sulfidic conditions, V will have relatively higher enrichment than Mo and U, whereas, in a sulfidic or
euxinic condition, Mo and U will undergo higher enrichment than V. PCA suggested that the V enrichment
is likely to have some detrital influence as well. Thus, relatively high enrichment of V compared to Mo and
U does not always indicate a more reducing condition. Consequently, in this study, trends in variation in V
concentration were used for chemostratigraphic correlation purposes, but absolute concentration was not
used to infer the redox conditions of the different chemostratigraphic units. In contrast, Mo and U
concentrations were used for both chemostratigraphic correlations and to infer the redox conditions in the
UBS.

Figure 3.20: Box and whisker plot showing the variation in the absolute concentration for the three redox-
sensitive trace elements: Mo, U, and V in UBS. Each plot comprises of all samples of UBS from all the
sixteen wells for which XRF data was collected in this study.

Figure 3.21: (a) Crossplot of Mo vs. U showing a strong positive covariance between the two elements. (b)
Cross-plot between Fe and S showing a positive covariance which indicates that most of the Fe and S are
associated with pyrite. DOPT (Degree of Pyritization) lines in the crossplot are based on Algeo and Maynard
(2004). All the samples are color-coded by the three different chemostratigraphic units of UBS, which are
described in Section 3.4.5.

80
Figure 3.22: (a) Cross-plot of Mo vs. V showing a weak covariance between the two elements. The data
points from three different clusters which are indicated in by encirclements. (b) Cross-plot of U vs. V
showing a weak correlation between the two elements. The data points from three different clusters are
indicated by encirclements. All the samples are color-coded by the three different chemostratigraphic units
of UBS, which are described in Section 3.4.5.

XRD-XRF comparison

This section illustrates the effectiveness of using major element (Si, Al, K, and Ca) concentrations
obtained from ED-XRF measurements as a proxy for mineral content in the context of the UBS. For the
Middle Bakken Member, Jin (2014) observed correspondence between Well B XRF data with the
QEMSCAN derived mineralogy from Kowalski (2010). A side-by-side profile view of XRD mineralogy
and elemental data from XRF in Figure 3.23 qualitatively compares the elements and the dominant minerals
composed of these elements. The profile view of the XRD-derived mineralogy and ED-XRF elemental data
for the UBS in wells LT and BF showed a definite correspondence between Si and quartz; Al and K with
clay; and Ca with dolomite. For example, Figure 3.23a shows that for well LT at the 10389-10390ft interval,
both Si and quartz content was high (35-53 wt.%); whereas in the 10383-10385ft interval, both Si and
quartz content was low (14-24 wt.%). Similar trends for Si were observed in well BF and are shown in
Figure 3.23b; Si was therefore assumed to be primarily associated with quartz in the UBS. In addition, the
10389-10390ft interval of well LT had low Al, K, and clay content (14-33 wt.%). This contrasts to the
10383-10385ft interval, which had higher Al and K content, and corresponded to a high clay content of 44-
50 wt.%. As K-felspar remained constant over the entire section of the UBS in well LT, Al and K content
variations can be attributed to the changes in the amount of clay. Moreover, at the 10386-10390ft interval
in well LT, relatively less Ca content was observed as compared to that in the upper 10383-10386ft interval.
Variation in Ca concentration correlates with the variation in carbonate content, especially dolomite. For
example, the lower interval has dolomite content of 6 wt.%, while in the upper interval, dolomite content
increased to 8-10 wt.%.

81
Figure 3.23: Depth profiles comparing the XRF data of major elements and XRD mineralogy of UBS in (a)
well LT, and (b) well BF.

82
 Elemental chemostratigraphic correlation

Based on the trends of the elemental proxies and lithofacies distribution, two chemostratigraphic
units, Unit-1 and Unit-2, were identified in the UBS. Unit-1 is the stratigraphically lower interval of the
UBS, while Unit-2 is located at its upper intervals. Unit-1 was further subdivided into the stratigraphically
lower sub-unit 1a and the upper sub-unit 1b. Table 3.2 lists differences between Units 1 and 2, as well as
the signature of the boundary marker.

Table 3.3 lists differences between sub-units 1a and 1b and their boundary marker signature. These
differentiating signatures of the units and sub-units are further elaborated through examples of the
chemostratigraphic profiles for wells R, RT, and G (Figure 3.24, Figure 3.25 and Figure 3.26). Profiles for
the other wells are included in Appendix D.

Table 3.2: Characteristic differences in terms of the biogenic, detrital, grain-size proxy, redox proxy, and
facies between Unit-1 and Unit-2 are listed. The signature of the boundary marker is also specified.
Biogenic, Detrital,
Unit Redox proxy Facies
Grainsize proxy

Dominated by massive to very finely

High Si and low Al, High Mo and U content compared to
1 laminated silt-bearing mudstone (F2)
K and Ti and Ca Unit-2.
followed by siliceous mudstone (F1)

Unit boundary marker: Characteristic peak especially in V immediately below the contact of 1 and 2
Low Si and high Al, Dominated by massive to very finely
V shows a characteristic blocky
K and Ti, and Ca and laminated silt-bearing mudstone (F2)
2 character, Mo and U less enriched
an upward increasing or Macrofossil-bearing silt-rich
than Unit-1
trend in Zr mudstone (F4)

Table 3.3: The characteristic differences between sub-unit 1a and 1b in terms of the biogenic, detrital, grain
size proxy, redox proxy and facies are listed. The signature of the unit boundary marker is also specified.
Biogenic, Detrital,
Unit Redox proxy Facies
Grainsize proxy
Al, K, and Ti in sub- Highest Mo and U content.
1a unit 1a slightly lower Characteristic peak most prominently
compared to sub-unit 1a in V just below 1a and 1b contact. Dominated by massive to very
Unit boundary marker: Characteristic peak most prominently in V finely laminated silt-bearing
immediately below the contact mudstone (F2) followed by siliceous
Al, K, and Ti in sub- mudstone (F1)
High Mo and U but comparatively
1b unit 1b slightly high
less than that in 1a.
compared to sub-unit 1a

83
Figure 3.24: Core description and chemostratigraphic profile with elemental trends of biogenic, detrital, carbonate and redox-proxying elements from
Rolf 1-20H core showing the chemostratigraphic sub-unit 1a, sub-unit 1b and Unit-2 of UBS. From left to right: Si-silicon; Al-aluminum; K-
potassium; Ti-Titanium; Zr-zirconium; Ca-calcium; Mo-molybdenum; U-uranium; V-vanadium. MFS-Maximum Flooding Surface.

84
Figure 3.25: Core description and chemostratigraphic profile with elemental trends of biogenic, detrital, carbonate and redox-proxying elements from
Roberts Trust 1-13H showing the chemostratigraphic sub-unit 1a, sub-unit 1b and Unit-2 of UBS. From left to right: Si-silicon; Al-aluminum; K-
potassium; Ti-Titanium; Zr-zirconium; Ca-calcium; Mo-molybdenum; U-uranium; V-vanadium; MFS-Maximum Flooding Surface.

85
Figure 3.26: Core description and chemostratigraphic profile with elemental trends of biogenic, detrital, carbonate and redox-proxying elements from
core Graham USA 1-15 showing the showing the chemostratigraphic sub-unit 1a, sub-unit 1b and Unit-2 of UBS. From left to right: Si-silicon; Al-
aluminum; K-potassium; Ti-Titanium; Zr-zirconium; Ca-calcium; Mo-molybdenum; U-uranium; V-vanadium. MFS-Maximum Flooding Surface.

86
 The chemostratigraphic profile used in this study consists of the core description, GR log, and
elemental concentration logs for Si, Al, K, Ti, Zr, Ca, Mo, U, and V. Figure 3.24 and Figure 3.25 show the
chemostratigraphic profiles for wells R and RT. The profiles indicate that Unit-1 had comparatively higher
Si, and lower Al, K, Ti, and Zr. Moreover, Unit-1 was characterized by an abundance of siliceous mudstone
(F1), which is consistent with the high Si observed in the unit. This difference in chemostratigraphic
characteristics between Units 1 and 2 manifests statistically through the clustering of data-points in two
distinct groups in all crossplots of Figure 3.18. Sub-unit 1a had comparatively lower Al and K than sub-
unit 1b, which is evident in the data-point clustering in the Al vs. K crossplot in Figure 3.18c. Furthermore,
Unit-1 is enriched in the redox-sensitive trace elements Mo, U, and V, as their concentration was higher
than average shale values (Section 3.3.3). This chemostratigraphic difference can be seen in the clustering
of two distinct groups of data-points in Figure 3.21a to 3.20c crossplots.

Additionally, Mo vs. V and U vs. V crossplots in Figure 3.21b and c also suggest that Unit-1 had
higher U and Mo and lower V compared to Unit-2, which had higher V and showed a characteristic blocky
signature. For sub-unit 1b, Mo and U were less enriched, while V was slightly more enriched compared to
sub-unit 1a. This is shown in the Mo vs. V and U vs. V crossplots in Figure 3.21b and c, in which they
appear to form different clusters with some overlap. The unit and sub-unit boundaries are marked by peaks
in the levels of redox-sensitive trace-elements. The V peak is the most predominant, appearing immediately
below the 1a and 1b contacts and the Unit 1 and 2 contacts (Figure 3.24 and Figure 3.25).

Regional Distribution of Chemostratigraphic Units

All three chemostratigraphic units and sub-units within the UBS were developed in all cores used
in this study. These units and sub-units of UBS are correlatable in wells across the study area from the
northeast to southwest (Figure 3.27) and from the southeast to northwest (Figure 3.28); however, a
considerable lateral variation in thickness was observed in the three different units. Table 3.4 lists the
thickness of the UBS and its different chemostratigraphic units and sub-units. Isochore maps for UBS, its
chemostratigraphic units and sub-units, which are shown in Figure 3.29, were prepared based on the
thickness information of Table 3.4. The UBS attains its maximum thickness in the basin center following a
north-northwest and south-southeast trend along the Nesson Anticline (Figure 3.29a). In this study, the
thickest UBS interval was encountered in cores RT and L, which are 21.2 ft and 21.5ft thick.

The total thickness of Unit-1 ranged from 4ft in well BY, which is located along the southwestern
margin of the basin at the Elm Coulee Field, to a maximum of 14 ft in well L, which is located at the center
of the basin (Figure 3.29b). The maximum thickness trend of Unit-1 followed the similar north-northwest
to south-southeast trend along the Nesson Anticline. Unit-1 loses its thickness along the southwestern

87
margin of the basin. The thickness of sub-unit 1a varied from 1.1ft in well BY, located at the southwestern
basin margin, to a maximum of 6.5ft in well C, located at the basin depocenter (Figure 3.29c). The thickness
of sub-unit 1a followed a north-northwest to south-southeast trend. Sub-unit 1b followed a similar north-
northwest and south-southeast trend as that of sub-unit 1a and attained a maximum thickness of 9ft in well
L, located in the southern part of the basin depocenter. The thickness of sub-unit 1b varied from 2ft in wells
LT, T, and G, located at the southwestern basin margin, to a maximum of 9ft in well L, located at the
southern end of the basin center (Figure 3.29d). Sub-unit 1b also loses its thickness along the southwestern
basin margin.

Unlike sub-units 1a and 1b, the thickness variation of Unit-2 was uniform with localized maxima
across the study area. Maximum thickness was developed along the southern end of the basin center in a
localized area, where Unit-2 attains its maximum thickness of 9.2ft in well RT (Figure 3.29e). Unit-2
pinches out in the southwestern basin margin such that it has a mere 1.2ft thick interval of well BN. This
indicates that Unit-2 was most likely not deposited further south towards the proximal end of the basin.

Table 3.4: Thickness of different chemostratigraphic sub-units and units of UBS in all the eighteen wells
used in this study. Well Abv- Abbreviation for well names.
Well Well Abv Sub-unit 1a Sub-unit 1b Unit 1 Unit-2 Total
(ft) (ft) (ft) (ft) (ft)
Miller 34X-9 M 2.5 7 9.5 3.5 13
Jorgensen 1-15H J 3 3 6 3.5 9.5
Saratoga 12-1 S 1.5 5 6.5 5 11.5
Graham USA 1-15 G 2.9 2.3 5.2 1.7 6.9
Bullwinkle Yahoo BY 1.1 3 4.1 1.7 5.8
Roberts Trust 1-13H RT 4.5 7.5 12 9.2 21.2
BN 1-23H BN 3 3 6 1.2 7.2
5-1 Thompson Unit TT 3 4 7 1.3 8.3
Braaflat 11-11H BF 3 4 7 4.5 11.5
Sidonia 1-06H S 4.8 5.9 10.7 5.3 16
Rasmussen 1-21-16H RM 4.5 5 9.5 2.9 12.4
Rolf 1-20H R 6 7 13 5 18
Charlotte 1-22H C 6.5 5.5 12 4.3 16.3
Linseth 13-12 L 5 9 14 7.5 21.5
Titan F-WP 32-14-H T 3.5 2 5.5 1.75 7.25
Round Prairie 1-17H RP 3.5 4 7.5 3 10.5
Wayjetta 44-0311H W 4 5 9 5 14
RR Lonetree Edna LT 2.5 2 4.5 2.5 7

88
Figure 3.27: NE-SW chemostratigraphic correlation between wells RR Lonetree Edna, Round Prarie 1-17H, Rolf 1-20H, Sidonia 1-06H and Jorgensen
1-15H.

89
Figure 3.28: NW-SE chemostratigraphic correlation between wells Rasmussen 1-21-16H, Charlotte 1-22H, Roberts Trust 1-13H, and Miller 34X-9.

90
Figure 3.29: (a) Isopach map of the UBS showing north-northwest to south-southeast trend of maximum
thickness in the basin depocenter. (b) Map for sub-unit 1a showing a north-northwest to south-southeast
trend of maximum thickness in the basin depocenter. Maximum thickness for sub-unit 1a is encountered in
well C (6.5ft). (c) Map of chemostratigraphic sub-unit 1b showing a north-northwest to south-southeast
trend of maximum thickness in the basin depocenter. Maximum thickness for sub-unit 1b is encountered in
well L (9ft). (d) Map of chemostratigraphic Unit-2 showing the development of maximum thickness around
wells L and RT in the southern part of the basin depocenter. Unit-2 loses its thickness drastically along the
southern basin margin. Maximum thickness for Unit-2 is encountered in well RT (9.2ft). An inset map in
each figure showing the isopach map of the entire UBS in US-portion of the Williston Basin and the shaded
box in the inset map represents the study area. Locations of wells are shown by black dots. Abbreviations
of well names used in the maps are listed in Table 3.1.

91
 Distribution of Facies in Chemostratigraphic Units

The contact between the UBS and the Middle Bakken Member, which is essentially the base of
sub-unit 1a, was found to be either gradational or sharp. Phosphate and fossil-bearing mudstone (F6) was
commonly present in sub-unit 1a when the contact was sharp, indicating a transgressive lag deposit. These
transgressive lag deposits at the basal contact of sub-unit 1a were usually 10-20 mm in thickness and were
found in both the basin margin (e.g., Well BN, Figure 3.16b) and the depocenter (e.g., Well R, Figure
3.16c); therefore, the presence of transgressive lag deposits at the basal contact of sub-unit 1a reflects the
local topography, rather than the proximity to the basin margin. The transgressive lag beds are usually
thicker in the southern margin of the basin compared to other parts of the basin, which probably indicates
that the area is located more proximal towards the basin margin.

This is followed by a 0.5-2ft interval in sub-unit 1a, which is dominated by massive to finely
laminated silt-bearing mudstone (F2) with intercalations of sand and clay-bearing silt-rich mudstone (F3).
Finer grain size in facies F3 was observed in the basin center compared to that in the basin margin, which
implies that storm-induced bottom currents deposited coarser silt in proximal parts of the basin, while finer
silt was deposited in the distal areas. Moving up in the section, sub-unit 1a was dominated by massive to
finely laminated silt-bearing mudstone (F2) with intercalations of siliceous mudstone (F1). The frequency
of siliceous mudstone (F1) beds varied laterally and was not laterally correlated, as shown in the correlation
profiles in Figure 3.30 and Figure 3.31. There was no visual difference between sub-unit 1a and 1b with
respect to the distribution of dominant facies type.

Sub-unit 1b was also dominated by finely laminated silt-bearing mudstone (F2) with intercalations
of siliceous mudstone (F1); however, sand and clay-bearing silt-rich mudstone (F3) and phosphate clasts
and fossil-bearing mudstone (F6) were only present within 2ft from the base of sub-unit 1a and were not
present in sub-unit 1b. The presence of facies F3 along with facies F2 at the base of sub-unit 1a, which is
followed by predominantly facies F2 and facies F1, indicates a change from a combination of bedload
transport processes by storm-induced bottom currents and suspension sedimentation at the base of sub-unit
1a to dominantly suspension sedimentation in sub-unit 1b. This decrease in storm-induced current activity
likely reflects a deepening of the basing due to relative sea level rise.

Unit-2 had an overall massive appearance and was distinguished by the absence of siliceous
mudstone (F1). This absence of facies F1 is also reflected in the low Si content in the chemostratigraphic
profiles (Figure 3.24). As shown in Figure 3.30 and Figure 3.31, Unit-2 is dominated by massive to faintly
laminated silt-bearing mudstone (F2, such as in Well S), or by macrofossil-bearing silt-rich mudstone (F4,
such as in well L), or it may have both facies present in variable proportions (such as in Well RM). Facies
F4 was primarily developed in a small area along the southern portion of the basin depocenter. As shown

92
in Figure 3.25, for Well RT, if Unit-2 is dominated by facies F4, the chemostratigraphic profile shows high
Ca, while Mo decreases to almost detrital background values (1-5 ppm; Zheng et al., 2000). On the western
side of the basin depocenter in wells C and RM, Unit-2 had a greater proportion of massive to faintly
laminated silt-bearing mudstone (F2) compared to macrofossil-bearing silt-rich mudstone (F4). Moreover,
Mo enrichment was moderate in Unit-2 (Appendix D, wells RM and C). Along the eastern side of the basin
depocenter, Unit-2 was mostly dominated by massive to faintly laminated silt-bearing mudstone (F2) and
had very high Mo and U concentration (Appendix-D, e.g., Well S). As shown in Figure 3.30 and Figure
3.31, in the southern margin of the basin, Unit-2 thinned out and mainly consisted of massive to faintly
laminated silt-bearing mudstone (F2) with intercalations of sand and clay-bearing silt-rich mudstone (F3),
which indicated that storm-induced bottom current activity was more pronounced in the proximal part of
the basin.

Phosphate and fossil-rich mudstone (F6) was present within the top 1-1.5ft of Unit-2 in many wells,
such as wells S, R and LT (Figure 3.30), which probably indicates that the relative sea level has
progressively descended, which increased the influence of storm-induced bottom water currents. The
contact of the UBS with the Lodgepole Formation marked the upper limit of Unit-2 and was characterized
by the presence of burrow-mottled silt-bearing mudstone (F5) (Figure 3.15). Occasionally, soft-sediment
deformation was also present at the boundary.

Gamma Ray Log and TOC Trends

Figure 3.32 shows the GR log and TOC profile for wells RT, BF, and S. As shown in Figure 3.32,
GR values are highest at the basal 2-4ft interval of the UBS; thereafter, it has a general decreasing upward
trend for the next 2-8ft, which is followed by another subtle peak in the upper 2-4ft interval. GR typically
attained the highest value in sub-unit 1a, which was around 400 units more than that of sub-unit1b. The
minimum value for GR was attained within the 2-5ft interval of the lower part of Unit-2. Temporal variation
in TOC was insignificant in wells BF and S, which did not show any relation to GR log variations. In well
R, TOC and GR values generally covary. For example, the TOC and GR logs showed an upward decreasing
trend and two intermittent peaks in the middle and top of the UBS.

93
Figure 3.30: Correlation is showing the lateral and vertical facies distribution in the three different chemostratigraphic units. The boundaries of the
three units marked in this correlation are picked from the chemostratigraphic profiles.

94
Figure 3.31: Correlation is showing the lateral and vertical facies distribution in the three different chemostratigraphic units. The boundaries of the
three units marked in this correlation are picked from the chemostratigraphic profiles of the wells.

95
Figure 3.32: Vertical profile of gamma ray (GR) log and SRA derived total organic carbon (TOC) for UBS
for wells: RT, BF, and S showing the presence of high GR peak near the base and top of UBS and an upward
decreasing trend in the middle of UBS interval in all the three wells. In well RT, TOC profile mimics GR
trend, whereas in the other two wells variation in TOC is insignificant and TOC profile does not follow
GR.

3.5 Discussion

This section integrates the results and their interpretation of the previous section to develop an
understanding of depositional conditions such as detrital sediment influx, sedimentation rate, depth of the
basin, and the redox conditions which prevailed during the deposition of the UBS in Williston Basin. This
section also discusses the likely position of the maximum flooding surface for the UBS and the associated
sequence stratigraphic implications. Finally, the chemostratigraphic units of the UBS are ranked for their
shale resource potential, primarily based on their mineralogical composition, natural fractures, and
brittleness.

96
 Detrital Sediment Influx

The Bakken Formation was deposited in the Late Devonian-Early Mississippian period when the
Williston Basin was located near the equator. In this phase, the basin is believed to have been under the
influence of the easterly trade wind system (Parrish, 1982; Golonka et al., 1994; Smith and Bustin, 1998).
Smith and Bustin (1998) proposed that the detrital fine-grained sediments in the UBS were mostly eolian
input from the surrounding Lauransian land mass to the east and north-east. The authors suggested that
these sediments were transported offshore by the prevailing east-west trade winds and were deposited by
suspension sedimentation. This interpretation of the sediment influx was probably based on two lines of
evidence: (i) the absence of prograding clinoform along the margin of the Williston Basin, and (ii) present-
day Saharan dust storms are reported to carry eolian material for 1000’s of miles offshore in the Atlantic
ocean (Middleton and Goudie, 2001). This study agrees with the eolian influx for the UBS deposition;
however, the findings of this study suggest that the sand and clay-bearing silt-rich mudstone (F3) was
deposited by weak bottom water currents as distal tempestites. Facies F3 showed a decrease in grain size at
the basin center (e.g., Well RT) compared to that in the southern basin margin (e.g., Well LT). This grain
size decrease suggests that apart from the northern and northeastern sediment influx, some portion of the
detrital silt and phosphatic clasts in facies F3 were probably derived from the southern basin margin.
Nonetheless, facies F3 is mostly present within 2ft from the bottom of Unit-1 and the top of Unit-2 in
proximal parts, and represents less than 5% of the entire UBS, suggesting that influx from the southern
basin margin was minor. Unit-2 had a higher concentration of detrital proxying elements Al, K and Ti
compared to those in Unit-1 in all wells, which indicates that Unit-2 has more detrital clay content. The
higher clay content in massive to faintly laminated silt-bearing mudstone (F2) and macrofossil-bearing silt-
rich mudstone (F4) in Unit-2 compared to Unit-1 is due to the dilution of detrital clay by biogenic silica in
Unit-1, rather than by any actual increase in the influx of detrital clay during the deposition of Unit-2. This
interpretation is based on the petrographic analysis which suggests the selective presence of radiolarians in
facies F2 in Unit-1 and their complete absence in the massive to faintly laminated silt-bearing mudstone
(F2) and macrofossil-bearing silt-rich mudstone (F4) facies in Unit-2.

Zr/Rb ratio is used in this study as a proxy for the variation of the silt-size grain content in the
chemostratigraphic units (Section 3.3.2). Lateral variation in the Zr/Rb ratio in Unit-2 suggests that during
the deposition of Unit-2, apart from eolian input from the northeast, at least a fraction of the detrital silt was
locally supplied from the southern margin of the basin. High silt content in macrofossil-bearing silt-rich
mudstone (F4) was predominantly present in Unit-2, especially in the southern part of the basin depocenter
wells (e.g., well RT). Unit-1 was dominated by massive to finely laminated silt-bearing mudstone (F2)
followed by siliceous mudstone (F1) in most locations. The contrast in silt content between Unit-1 and

97
Unit-2 in well RT is shown in Figure 3.25 by an increase in both Zr and the Zr/Rb ratio in Unit-2 in
comparison to Unit-1. A similar trend of high Zr and Zr/Rb ratio in Unit-2 was also observed in well L;
therefore, an increase in Zr/Rb ratio generally reflects the higher influx of detrital silt during deposition of
Unit-2.

Where Unit-2 mainly consisted of dominantly massive to finely laminated silt-bearing mudstone
(F2), the Zr/Rb ratio did not vary significantly between the Units. This indicated that influx of detrital silt
remained same during the deposition of the Units in locations other than the southern part of the depocenter.
These contradicting observations about detrital silt influx during the deposition of Unit-2 suggest that
detrital silt was derived from multiple sources during the deposition of Unit-2, and result in lateral variation
in facies. The increase in the Zr/Rb ratio and presence of facies F4 in Unit-2 which develops more towards
the southern part of the depocenter in wells RT and L indicates that the additional detrital siliciclastic silt
and detrital dolomite were sourced from the southern basin margin.

An increase in detrital silt content during the deposition of Unit-2 was localized and is associated
with a facies change, which suggests that the influx of detrital sediment from the southern margin of the
basin was likely carried to the basin interior by weak bottom water currents. A universally low Zr content
in Unit-1 in comparison to that in Unit-2, however, has resulted from a dilution effect of biogenic silica
associated with radiolarians in Unit-1. The increase in Zr content in Unit-2 was probably due to the absence
of a dilution effect by biogenic silica in this unit. This should not be interpreted as an actual increase in the
influx of detrital silt influx during the deposition of Unit-2. However, in wells G (Figure 3.26), T, LT, and
BY (Appendix D), which are located along the southern basin margin, Zr content in Unit-1 was similar to
that in Unit-2 and was comparatively higher than other parts of the basin. This probably indicates the
presence of silt and sand-sized sediments associated with the facies F3, which is found more along the basin
margin area.

Sedimentation Rate

Previous studies by Hayes (1985), Smith and Bustin (1996, 1998) and Longman et al., (2014)
suggest that the UBS was deposited in an offshore environment by continuous suspension settling of fine-
grained sediments and organic matter at a slow rate in stagnant and anoxic bottom water of the basin. Based
on this model of deposition, the sedimentation rate of the UBS was proposed by Smith and Bustin (1998)
as 2 mm/1000 years. Longman et al., (2014) proposed a sedimentation rate of 0.7mm/1000 years based on
the total thickness of the UBS in well L adjusted to the geochronologic timescale given by Haq and Schutter
(2008); however, there is controversy regarding the time span for the deposition of the UBS due to
insufficient radiometric age control of the Bakken Formation, which would aid in the accurate

98
determination of the average sedimentation rate. Longman et al., (2014) suggested that the UBS was
deposited over a period of 7MY, while Sonnenberg (2015) suggested that the UBS was deposited in a period
of 1MY.

Petrographic characterization of the UBS in this study shows that sedimentary processes were not
constant, and instead varied through time during the deposition of the UBS; therefore, sedimentation rate,
which is dependent on the depositional processes, is also expected to vary for each facies type in the UBS.
Massive radiolarian-dominated mudstone (F1a) was deposited as a result of episodic radiolarian blooms by
suspension sedimentation. Sand and clay-bearing silt-rich mudstone (F3) and phosphatic clast and fossil-
rich mudstone (F6), which represent distal tempestites, were deposited by bedload transport processes.
Although these three facies (F1a, F3, and F6) were deposited by different physical processes, they all
represent episodic sedimentation at a comparatively faster rate. In contrast, the massive to finely laminated
silt-bearing mudstone (F2) was deposited predominantly by suspension sedimentation of organic matter,
organominerallic aggregates, fecal pellets and detrital silt and clay. Macquaker et al., (2010) suggested that
organic matter is generally deposited as organominerallic aggregates during episodic algal blooms as
marine snow, and such organic-rich mudstones are stacked into mm-thin beds with sharp bases and are
overlain by silt lags; however, organic-rich facies F2 in the UBS have gradational bed boundaries and
laminasets, and silt lags were also not observed in this facies. This probably indicates that the
organominerallic aggregates and fecal pellets were delivered to the basin continuously at a slow rate with
occasional episodic algal blooms. Recent studies have shown that organominerallic aggregates or marine
snow, phytoplankton, and fecal pellets of zooplanktons settle through the water column at a faster rate than
individual grains (Shanks, 2002; Turner, 2002, 2015). Organominerallic aggregates and phytoplanktons
have sinking rates in the range of 16-368 m/day and 100-150 m/day, respectively (Turner, 2002). This
suggests that during the deposition of facies F2, the organic matter could have settled as marine snow at a
comparatively faster rate.

Depth of Basin

Smith and Bustin (1998) proposed that the UBS was deposited below the storm wave base at a
water depth of more than 200 m based on the laminated fabric of the organic-rich shales, the absence of
storm deposits, and lack of bioturbation and body fossils. The suggested water depth by Smith and Bustin
(1998) would demand repeated drastic changes in sea level over very short periods of time. Although the
overall long-term, cumulative sea level rise could be as much as 250 m, individual third-order changes in
sea level occurring over ~0.5 to 6 million years and rarely exceeded 150 m (Haq and Schutter, 2008).
Another point of concern with such great water depths is the lack of clinoforms in the entire Bakken
Formation, which should be expected on a slope of over 200m. Moreover, the depth of 200 m is a common

99
estimate for storm wavebase (Boggs, 1987). Recent studies by Peters and Loss (2012) have demonstrated
that the wave base rarely exceeds 150 m in the Atlantic Ocean, and is even shallower, generally less than
100m, along with the Gulf Coast. Based on these observations, Scott et al., (2017) proposed a maximum
water depth of 100-150 m during the deposition of the UBS.

Facies F3 and F6 of the UBS were interpreted to be deposited as distal tempestites by storm-induced
bottom currents. This indicates that at least occasionally during the initial stages of transgression, the depth
of the basin was near the storm wave base along the basin margin. In semi-enclosed epicontinental basins
such as the Williston Basin, storm wave base is not deeper than a few tens of meters, down to a maximum
of about 50 m (Shaw, 1964; Aigner and Reineck, 1982; Peters and Loss, 2012; Schieber, 2016). So, at least
during the initial deposition of the UBS, the maximum water depth along the basin margin of the Williston
Basin was probably about 50m. The presence of radiolarians, however, indicates accumulation of sediments
in deep basinal offshore settings at a water depth of more than 50m (Stasiuk and Fowler, 2004). Therefore,
the water depth was more than 50m during the deposition of sub-units 1a and 1b, which are characterized
by an abundance of radiolarian beds associated with facies F1. The depth of the basin during the deposition
of UBS was likely never higher than 200 m.

Redox conditions and bioturbation

There are two schools of thought regarding the paleo-redox conditions that prevailed during the
UBS deposition. The first of these two interpretations is more widely accepted, and suggests that the UBS
was deposited mainly by suspension sedimentation in a restricted basin in an anoxic-euxinic bottom water
condition (Webster, 1984; Smith and Bustin, 1996, 1998; Sonnenberg and Pramudito, 2009; Longman et
al., 2014). This interpretation is largely based on the following observations, 1) the presence of planar, thin
laminations within the shales resulting in deposition below storm wave base in a tranquil condition; 2) the
presence of planktonic algal spores (Tasmanites sp.), fish remains, cephalopods, ostracods, conodonts, and
inarticulate brachiopods representing pelagic organisms or rafted skeletal debris derived from the
oxygenated surface water; and 3) the rare presence of trace fossils or body fossils along with the presence
of high amounts of organic and pyritic material, indicating benthic anoxia (Webster, 1984; Smith and
Bustin, 1996; Longman et al., 2014). This interpretation was further supported by high enrichment of redox-
sensitive trace elements like Mo, U, V, and Ni in the UBS, which in turn indicated that the UBS was
deposited in anoxic to euxinic conditions (Kocman, 2014).

The second school of thought, proposed by Egenhoff and Fishman (2013), is based on petrographic
study and proposes that the UBS was deposited in dysoxic conditions. Egenhoff and Fishman (2013)
reported the presence of vertical microburrows and sedimentary structures, indicating bedload transport

100
processes in the UBS. Based on these observations, the authors interpreted that the UBS was predominantly
deposited in a dynamic environment with suboxic to dysoxic bottom water conditions. Schieber (2014)
argued that the features identified as vertical microburrows in the UBS by Egenhoff and Fishman (2013)
were in fact artifacts in the petrographic thin sections.

My findings suggest that during the deposition of sub-units 1a and 1b, bottom water conditions
were predominantly euxinic. This interpretation is based on the observation of the lack of bioturbation and
body fossils, the presence of dominantly planktonic fecal pellets and abundance of pyrite framboids in
massive to finely laminated silt-bearing mudstone (F2), and the predominant suspension sedimentation of
fine-grained sediments and radiolarians with only occasional weak bottom water current activity. Moreover,
euxinic conditions in bottom water are supported by the high enrichment of Mo and U found in sub-units
1a and 1b. Scott and Lyons (2012) suggested that Mo concentration of more than 100 ppm in shales
indicates that a persistent euxinia existed in the water column. More than 85% of all sub-unit 1a and 1b
samples used in this study (from all wells) were found to have more than 100ppm Mo (Figure 3.33a and b).
Moreover, 90% of the samples from sub-unit 1a and 70% of the samples from sub-unit 1b had DOPT values
greater than 0.75. This further supports that sub-units 1a and 1b were deposited under euxinic benthic
conditions, and represents deposition under strongly stratified water column based on the threshold values
of DOPT > 0.75 for euxinic conditions, as suggested by Raiswell et al.(1988), Hatch and Leventhal (1992),
and Rimmer et al. (2004). Sedimentological features and inorganic geochemical results, therefore, indicate
that sub-units 1a and 1b were deposited in predominantly euxinic bottom water conditions with occasional
storm-induced bottom current activity.

Figure 3.33: Histograms showing the statistical distribution of Mo concentration in various

chemostratigraphic units of UBS. a) For sub-unit 1a, more than 90% of the samples have Mo concentration
more than 100ppm; b) For sub-unit 1b, more than 80% of the samples have Mo concentration more than
100ppm; b) For Unit-2, more than 60% of the samples have Mo concentration more than 100ppm.

101
 Occasionally, smaller microscopic-scale Planolites burrows were observed in the clay-rich top
laminae of facies F3 along the southern basin margin (Figure 3.11b). Discrete occurrences of these burrows
indicate that after deposition of these beds by bedload transport processes, a slightly dysoxic bottom water
condition developed for a limited time, which allowed the diminutive fauna to colonize the clay-rich tops
of these mudstones. The high influx of organic matter, however, resulted in rapid exhaustion of the oxygen
content in the bottom water and the sediment-water interface soon became anoxic to euxinic. In sub-unit 1a
in well G, Mo concentrations were consistently between 25-100ppm along with low enrichment of U and
V. Scott and Lyons (2012) suggested that a Mo concentration between 25-100ppm implies intermittent
euxinic conditions. This indicates that in well G, which is located along the southern basin margin, benthic
conditions were less reducing compared to other parts of the basin.

In most of the cores, Mo and U concentration in sub-unit 1a was the highest among all three
chemostratigraphic units. Cores located in the central part of the basin, where the thickness of sub-unit 1b
was at its maximum, both Mo and U concentration displayed an upward decreasing trend while progressing
into sub-unit 1b. A decrease in the absolute concentration of Mo and U in sub-unit 1b suggests either a
decrease in the availability of the dissolved Mo or U ions within the water column or a decrease in the
supply of organic matter, which is the host-phase for these redox-sensitive trace elements. (Algeo and
Lyons, 2006; Algeo et al., 2007). A decrease in the availability of Mo and U ions would suggest an increase
in the degree of basin restriction due to a fall in relative sea level (Algeo and Lyons, 2006; Algeo et al.,
2007), during the deposition of sub-unit 1b; however, as shown in Figure 3.26, in well G, which is situated
close to the southern basin margin, sub-unit 1b had the highest Mo and U concentrations. A decrease in Mo
and U concentration due to basin restriction would have affected the margin of the basin more strongly;
therefore, Mo and U concentrations in sub-unit 1b in well G ideally would not have increased. This suggests
that during the deposition of sub-unit 1b, the availability of Mo ions in the water column did not decrease
due to increasing basin restriction, but rather the basin margin became more reducing probably due to a
relative sea level rise. A decrease in Mo and U concentration in sub-unit 1b in wells located in the basin
depocenter, therefore, suggests a decrease in the availability of the host-phase, i.e., organic matter.

Sub-unit 1a was characterized by strong enrichment in Mo concentration, and a DOPT > 0.75. This
indicates that bottom water conditions were strongly reducing, which likely resulted in the higher
preservation of organic matter. After the initial sea transgression, which brought nutrient-rich water, organic
productivity increased, resulting in the high amount of deposition of organic matter that rapidly depleted
the oxygen content at the bottom of the basin. This caused rapid development of euxinic conditions at the
sediment-water interface, but the surface water remained well oxygenated and sustained high organic
matter productivity, as evident from the presence of planktonic fecal pellets in massive to finely laminated

102
mudstone (F2) and the abundance of radiolarians in siliceous mudstone (F1). Egenhoff and Fishman (2013)
suggested that the UBS is completely bioturbated by vertical micro-burrows identified as Phycosiphon
incertum, but such burrows were not observed in any of the facies in sub-units 1a and 1b. Additionally,
Schieber (2014) suggested that these shales undergo high compaction, and it is highly unlikely that any
vertical burrow has remain uncompacted.

During the deposition of Unit-2, redox-conditions varied laterally across the basin from suboxic to
euxinic. Variability in redox condition was also present in the lateral variability in a given facies. There
was an overall decrease in Mo and U in all wells from both basin depocenter and basin margin (except wells
S and ST, located in the north-eastern part of the basin). About 45% of all samples from Unit-2 from all
wells had a Mo concentration of less than 100ppm, indicating an overall decrease in Mo concentration
regionally (Figure 3.33c). Scott and Lyons (2012) suggested that Mo concentrations between 25-100 ppm
indicate either or both of the following: i) intermittent euxinic conditions in the bottom water, ii) decrease
in availability of dissolved Mo ions in the water column. In wells R (Figure 3.25) and L, which are located
near the basin depocenter, Unit-2 was dominated by macrofossil-bearing silt-rich mudstone (F4). Mo and
U concentrations were very low in most of the Unit-2 intervals, such that the Mo concentrations were as
low as 2-3 ppm, equivalent to the average crustal values. Scott and Lyons (2012) suggested that a Mo
concentration lower than 25 ppm indicates that euxinic conditions are restricted only within the pore space,
while the sediment-water interface is in suboxic to anoxic conditions.

As discussed in Section 3.4.1, facies F4 was also characterized by the presence of agglutinated
benthic forams and comparatively low TOC (3-8 wt.%); therefore, a limited enrichment of Mo and U, along
with the presence of agglutinated benthic forams and low TOC, suggests that near the southern part of the
basin depocenter, the sediment-water interface was suboxic to anoxic during the deposition of Unit-2.
Development of suboxic conditions in basin depocenter suggests that the basin-wide lowering of Mo and
U concentrations during the deposition of Unit-2 was probably the result of less reducing conditions across
the basin. Because of this, in the northern and central parts of the basin where Unit-2 was dominated by F2
and Mo concentration was less than 100ppm, benthic conditions were only intermittently euxinic.
Furthermore, the overall lowering of reducing conditions is supported by low DOPT values for Unit-2
samples.

As shown in Figure 3.21d, DOPT values from about 70% of the samples from Unit-2 varied from
0.75 to 0.5. Raiswell et al., 1988; Hatch and Leventhal, 1992; Rimmer et al., 2004 suggested that when the
DOPT is less than 0.75 but greater than 0.42, it indicates the presence of anoxic to dysoxic conditions and
a less strongly stratified water column. Results from the multiple redox-proxies of Mo, U, and DOPT suggest
that during the deposition of Unit-2, the bottom water condition was less reducing compared to Unit-1 such

103
that in some areas it became suboxic. In wells S and ST (Appendix D), Mo concentration was persistently
high (>100 ppm), along with a high enrichment of U in Unit-2, suggesting that persistent euxinic conditions
were also present locally within the basin. The amounts of oxygen present within the water column along
the sediment-water interface during the deposition of Unit-2, therefore, varied laterally within the basin.

Presence of macroscale Planolites burrows and the development of burrow-mottled silt-bearing

mudstone (F5) at the contact of Unit-2 with the overlying Lodgepole Formation suggests that benthic
conditions became progressively oxygenated with the onset of deposition of the Lodgepole Formation. This
study demonstrates that redox conditions varied from strongly euxinic to suboxic both temporally and
regionally during the deposition of the UBS. Moreover, the degree of bioturbation in the UBS was most
likely overestimated by Egenhoff and Fishman (2013), and this has led to an incorrect interpretation that
the Williston Basin was predominantly dysoxic during the deposition of the UBS.

Sequence Stratigraphic Interpretation

Most of the previous studies on the sequence stratigraphic interpretation of the Bakken Formation
were focused on the Middle Bakken Member (Smith et al., 1995; Smith and Bustin, 2000; Kohlruss and
Nickel, 2009; Angulo and Buatois, 2012). Theloy (2014) compiled the previously proposed sequence
stratigraphic models for the Bakken Formation and concluded that in most models, the upper part of the
Middle Bakken shows a fining upward lithology; therefore, it represents the beginning of a major
transgression. Most previous studies included the upper part of the Middle Bakken and the entire UBS
within the transgressive system tract (TST), while the overlying Lodgepole Formation was included within
the highstand system tract (HST); however, one of the major issues with constructing the sequence
stratigraphic framework in the organic-rich UBS is identifying the maximum flooding surface (MFS). This
is because, unlike many other mudstones, MFS in the UBS is not marked by any distinct, recognizable
condensed sections distinguished by the presence of phosphatic hardgrounds, regionally correlatable
phosphatic granule layers, preferential cementation, sediment starvation, or unusual enrichment of organic
content (Macquaker and Gawthorpe, 1993; Algeo et al., 2004; Macquaker et al., 2007).

Smith and Bustin (2000) suggested that the upper contact of the UBS should be considered as the
MFS since sedimentation of fine-grained organic-rich sediment continued at a slow rate until sea level
stopped rising and maximum water depth was reached. Deposition of the UBS ended with the establishment
of a sea level highstand, and the associated changes in water circulation and sedimentation patterns resulted
in the deposition of the overlying Lodgepole Formation. This interpretation, however, is inconsistent with
the following three major findings of this study, i) basin-wide disappearance of radiolarians in Unit-2, ii)

104
overall lowering of the reducing condition during the deposition of Unit-2, and iii) deposition of
macrofossil-bearing silt-rich mudstone (F4) in suboxic conditions at the basin depocenter.

As discussed in Section 3.4.1, radiolarians are deep-water planktonic organisms and are sensitive
to changes in relative sea level that cause changes in surface-water temperature, salinity, and nutrient supply
(De Wever and Baudin, 1996; Wang et al., 2006). The disappearance of radiolarians during the deposition
of Unit-2, therefore, most likely indicates a relative sea level fall. The overall lowering of reducing
conditions and the development of suboxic conditions at certain areas near the basin depocenter suggests
that due to relative sea-level fall during the deposition of Unit-2, the stratified water column was disturbed,
which led to the cessation of widespread euxinic condition across the basin. Moreover, a high Zr/Rb ratio
in sub-unit 2 in areas where macrofossil-bearing silt-rich mudstone (F4) was deposited suggests that influx
of detrital silt content increased during the deposition of Unit-2. This, in turn, suggests that more area along
the basin margin became exposed due to falling sea levels, which resulted in more weathering of silt
sediments. These evidence collectively suggest that relative sea level started falling with the deposition of
Unit-2; thus, the contact between the UBS with the Lodgepole Formation is unlikely to be the MFS.

Jin (2014) suggested that the MFS corresponds to the most reducing conditions and can be
identified in organic-rich shales from the maximum V and Ni enrichment near the upper contact of the
UBS. Based on this theory, Jin (2014) identified a sharp peak in V concentration as the MFS value near the
upper contact of the UBS in well BF (Appendix D). Although a sharp peak in V concentration was observed
near the upper contact of the UBS in few wells such as BF and RT, in most wells used in this study, no such
peak was observed. V showed a blocky character in Unit-2, and because of this observation, the sharp peak
in V concentration observed by Jin (2014) cannot be regionally correlated. Moreover, as discussed in
Section 3.3.3, V concentration in the UBS is not suitable to determine the maximum reducing condition
because i) V has a detrital influence in the UBS, and ii) enrichment of V takes place by a two-step reduction
process with separate carrier-phases, which may lead to the high enrichment of V at a relatively rapid rate
even in less reducing conditions (Algeo and Maynard, 2004; Tribovillard et al., 2006; März et al., 2008).
Lewan and Maynard (1982) also suggested that a high V enrichment can also suggest that anoxic or euxinic
conditions existed only within the pore water, while the sediment-water interface and water column were
less reducing. Since Unit-2 was deposited during a relative sea-level fall, this indicates that the MFS is
probably located below Unit-2.

The contact between the Middle Bakken and Unit-1 of the UBS represents a transgressive surface
and is marked by the presence of 10-20 mm thick phosphatic lag deposits (F6) in some places. Due to the
lack of conodont data from the Middle Bakken, it is not clear whether the transgressive surface represents
a significant hiatus in deposition (Figure 3.2). According to Wignall (1991), the basal part of transgressive

105
black shales is characterized by high GR and TOC values. Passey et al. (2010) suggested that in a
platform/ramp setting like the Williston Basin, the maximum TOC value occurs in the basal TST and
decreases stepwise to background values at the MFS. In almost all wells, the highest GR values were
observed in sub-unit 1a of the UBS, which progressively decreases upward in sub-unit 1b and peaks again
in Unit-2. In well R, TOC values mimic GR logs such that highest GR values are observed along with the
highest TOC at sub-unit 1a; therefore, sub-unit 1a is part of TST, which is consistent with the proposition
of Wignall (1991) and Passey et al. (2010).

The presence of sand and clay-bearing silt-rich mudstone (F3), along with massive and finely-
laminated silt-bearing mudstone (F2), at the base of sub-unit 1a, which is followed by predominantly facies
F2 and siliceous mudstone (F1), indicates a change from a combination of bedload transport processes and
suspension sedimentation at the base of sub-unit 1a to dominantly suspension sedimentation with occasional
weak bottom water current activity in sub-unit 1b. Moreover, in well G, located close to the south-western
margin of the basin, the highest Mo and U concentration was observed in sub-unit 1b. This indicates that
due to rising sea levels, maximum reducing conditions occurred during the deposition of sub-unit 1b,
implying a gradual rise in the relative sea level during the deposition of sub-unit 1a to sub-unit 1b; and
therefore, Unit-1 represents a transgressive cycle. The characteristic peaks in redox-sensitive trace elements
represent a flooding surface, such as in Figure 3.25 and Figure 3.26, especially in V, which is present
slightly below the contact boundary between sub-unit 1b and sub-unit 1a.

The basin-wide disappearance of radiolarians, an overall decrease in reducing conditions evident

from low Mo and U concentrations and DOPT values less than 0.75, localized development of suboxic
conditions, and an increase in the influx of detrital silt during the deposition of Unit-2 all suggest that Unit-
2 was deposited during relative sea-level fall and represents a regressive cycle. The MFS for the UBS was
thus most likely located near or below the contact between sub-unit 1b and Unit-2. A characteristic peak in
redox-sensitive trace elements, especially in V, represents the MFS in the UBS. A carbonate concretion
was also observed in well J (Figure 3.30) below the contact between Unit-1 and Unit-2. Concretionary
carbonates generally precipitate in response to either complete breaks in sedimentation or to periods where
sedimentation rates were very low (Macquaker and Gawthorpe, 1993); therefore, the presence of the
carbonate concretion in well J probably indicates very low sedimentation rate, which in turn further supports
that the MFS in the UBS is located just below the contact of Unit-1 and Unit-2. On the basis of the MFS
identified in this study, Unit-1 is interpreted to be part of the TST, while Unit-2 and the overlying Lodgepole
Formation is interpreted to be part of HST.

106
 Mineralogy Based Resource Potential of UBS Chemostratigraphic Units

In this section, the chemostratigraphic units of the UBS are ranked for their shale resource potential
based on their mineralogy, brittleness, and occurrence of natural fractures, which are among the key
indicators for a successful liquid-rich shale play (e.g., Passey et al., 2010; Bohacs et al., 2013). Based on
biogenic and detrital proxying elements, it was shown in Section 3.4.5 that the chemostratigraphic units of
the UBS have different biogenic silica, clay, carbonate, and detrital silt contents. Silica associated with
siliceous mudstone (facies F1) as radiolarians were attributed as the reason behind the high Si content in
Unit-1 in comparison to that in Unit-2. Moreover, a fraction of silica in facies F2 is likely to be authigenic
in nature (Schieber et al., 2000). Quartz associated with biogenic silica, which is present in Unit-1 only, is
more brittle compared to detrital quartz because recrystallized biogenic and authigenic silica creates a rigid
high modulus framework (Jarvie et al., 2007; Passey et al., 2010; Blood et al., 2013); therefore, Unit-1
should be more prone to natural fracturing, which is confirmed by the abundant presence of mineralized
ptygmatic natural fractures in Unit-1, especially in facies F1. Unit-1 was additionally characterized by
comparatively low detrital clay proxying elements Al, K, and Ti, which indicated lower clay content in
Unit-1. Lower clay content in shales decreases the likeliness of the adverse effects of clays with respect to
completion and production issues related to clay-swelling and fines migration (Passey et al., 2010; Blood
et al., 2013). High silica content, low clay fraction, and presence of mineralized natural fractures (Figure
3.5a) collectively indicate that Unit-1 is comparatively more brittle than Unit-2.

Brumsack (1989) suggested that 5*Al2O3 is an indicator for clay, SiO2 is an indicator for quartz,
and 2*CaO is an indicator for carbonate minerals in rock samples. Thus, a ternary plot between 5*Al2O3-
SiO2-2*CaO can show the relative concentration of clay, quartz, and carbonate in given samples. According
to Passey et al. (2010), in a quartz-clay-carbonate ternary diagram, currently producing shale plays
generally lie below the 50% clay line since shale-plays that contain greater than 50 wt.% quartz or carbonate
tend to be more brittle, and therefore are more fraccable.

In this study, the concentrations of Al2O3, SiO2, and CaO were calculated stochastically from the
absolute values of XRF data derived Al, Si, and Ca concentrations. Figure 3.34 shows a ternary diagram
between 5*Al2O3-SiO2-2*CaO for the different chemostratigraphic units of the UBS. The ternary diagram
shows that samples from Unit-2 always plotted above the 50% clay line, suggesting that Unit-2 is less brittle
and not a good target interval. Samples from sub-unit 1a and 1b, however, both plotted above and below
the 50% clay line. This indicates that the silica-rich intervals below the 50% clay line would be relatively
more brittle and therefore could be potential landing targets. This conclusion is further supported by the
XRD data available from five wells. Based on the XRD mineralogy, all Unit-2 samples plotted above the
50% clay line in the ternary diagram in Figure 3.34, whereas all samples from sub-units 1a and 1b plotted

107
below the clay line. The ternary diagram shows that few samples from sub-unit 1a, which plots close to the
quartz-rich corner, represent radiolarite or siliceous mudstone (F1). Based on the mineralogy and brittleness
properties, Unit-1a and 1b are better target intervals than Unit-2.

Figure 3.34: Ternary plot representing 5*Al2O3-SiO2-2*CaO relationship (after Brumsack, 1989) of the
chemostratigraphic sub-unit 1a, 1b, and Unit-2 of UBS. The three axes of (5*Al2O3), SiO2 and (2*CaO)
represents clays, quartz and/or biogenic silica, and total carbonates respectively. The black data points
represent XRD mineralogy from five wells, shape of the data points represents the three different
chemostratigraphic units. The ternary plot includes all the samples from the sixteen wells for which XRF
data has been collected in this study. All the data points are color-coded by the three different
chemostratigraphic units of UBS. The data points form mainly two clusters, which correspond to the two
main chemostratigraphic units: Unit-1 and Unit-2. Sub-unit 1b predominantly clusters below the 50% clay
line, whereas Unit-2 plots above the 50% clay line.

3.6 Conclusions

The main conclusions from this study are as follows:

• The UBS consists of six different lithofacies identified from core descriptions and petrographic
studies. These six lithofacies are: i) siliceous mudstone (F1); (ii) massive to finely laminated silt-
bearing mudstone (F2), (iii) sand and clay-bearing silt-rich mudstone (F3), (iv) macrofossil-bearing
silt-rich mudstone (F4), (v) burrow-mottled silt-bearing mudstone (F5), and (vi) phosphate and
fossil-rich mudstone (F6). Out of these six lithofacies, 95% of the UBS is composed of Facies F2
and F1.

108
 • Facies F1 and F2 were deposited by suspension sedimentation, whereas F3 and F6 were deposited
by bedload transport processes; however, due to bioturbation, the depositional processes for F4, F6
were not identifiable.
• Internal stratigraphic framework for the UBS was provided by identifying two laterally correlatable
chemostratigraphic units 1 and 2; Unit-1 was further subdivided into laterally correlated sub-unit
1b and sub-unit 1b. A total of three regionally correlatable chemostratigraphic packages were
identified in the UBS.
• For both sub-unit 1a and 1b, the maximum thickness followed a northeast and southwest trend in
the basin depocenter; for Unit-2, the thickness was localized and maximized near the southern part
of the depocenter.
• Influx of silt-size detrital sediments had multiple sources, especially during the deposition of Unit-
2. Along with eolian silt from the northeast, some detrital silt were also derived from the southern
basin margin.
• Redox conditions varied both temporally and laterally during the deposition of the UBS. Sub-units
1a and 1b, in most locations across the basin, were deposited in a persistently euxinic condition.
Unit-2 was deposited in a much less reducing condition, which varied from sub-oxic to
intermittently euxinic.
• The maximum flooding surface is located in sub-unit 1b below the Unit-2 and sub-unit 1b contact;
therefore, Unit-1 is part of the TST and Unit-2 is within the HST.
• Sub-units 1a and 1b have more biogenic and authigenic silica compared to Unit-2, which makes
them comparatively more brittle than Unit-2. As such, sub-units 1a and 1b will be better reservoir
targets compared to Unit-2.

3.7 Reference

Aigner, T., and H.-E. Reineck, 1982, Proximality trends in modern storm sands from the Helgoland Bight
(North Sea) and their implications for basin analysis.: Senckenbergiana Maritima, v. 14, no. 5–6, p.
183–215.
Algeo, T. J., and T. W. Lyons, 2006, Mo-total organic carbon covariation in modern anoxic marine
environments: Implications for analysis of paleoredox and paleohydrographic conditions:
Paleoceanography, v. 21, no. 1.
Algeo, T. J., T. W. Lyons, R. C. Blakey, and D. Jeffrey Over, 2007, Hydrographic conditions of the
Devono–Carboniferous North American Seaway inferred from sedimentary Mo–TOC relationships:
Palaeogeography, Palaeoclimatology, Palaeoecology, v. 256, no. 1, p. 204–230.
Algeo, T. J., and J. B. Maynard, 2004, Trace-element behavior and redox facies in core shales of Upper
Pennsylvanian Kansas-type cyclothems: Chemical Geology, v. 206, no. 3–4, p. 289–318.
Algeo, T. J., L. Schwark, and J. C. Hower, 2004, High-resolution geochemistry and sequence stratigraphy
of the Hushpuckney Shale (Swope Formation, eastern Kansas): Implications for climato-

109
 environmental dynamics of the Late Pennsylvanian Midcontinent Seaway: Chemical Geology, v. 206,
no. 3–4, p. 259–288.
Angulo, S., and L. A. Buatois, 2012, Integrating depositional models, ichnology, and sequence stratigraphy
in reservoir characterization: The middle member of the Devonian-Carboniferous Bakken Formation
of subsurface southeastern Saskatchewan revisited: AAPG bulletin, v. 96, no. 6, p. 1017–1043.
Arthur, M. A., and B. B. Sageman, 1994, Marine Black Shales: Depositional Mechanisms and
Environments of Ancient Deposits: Annual Review of Earth and Planetary Sciences, v. 22, no. 1, p.
499–551.
Blood, R., Lash, G. and Bridges, L. 2013. Biogenic silica in the Devonian Shale Succession of the
Appalachian Basin, USA. Search and Discovery Article #50864.
Boggs, S., 1987, Principles of Sedimentology and Stratigraphy: Merrill Publishing Company.
Bohacs, K. M., Q. R. Passey, M. Rudnicki, W. L. Esch, and O. R. Lazar, 2013, The Spectrum of Fine-
Grained Reservoirs from ‘Shale Gas’ to ‘Shale Oil’/ Tight Liquids: Essential Attributes, Key Controls,
Practical Characterization: IPTC 2013: International Petroleum Technology Conference 16676, p. 1–
16.
Brumsack, H.-J., 1989, Geochemistry of recent TOC-rich sediments from the Gulf of California and the
Black Sea: Geologische Rundschau, v. 78, no. 3, p. 851–882.
Calvert, S. E., 1976, The mineralogy and geochemistry of near-shore sediments: Chemical oceanography,
v. 6, p. 187–280.
Calvert, S. E., and T. F. Pedersen, 2007, Chapter Fourteen Elemental Proxies for Paleoclimatic and
Palaeoceanographic Variability in Marine Sediments: Interpretation and Application: Developments
in Marine Geology, v. 1, no. 7, p. 567–644.
Carlisle, J., L. Dryff, M. Fryt, J. Artindale, and H. von Der Dick, 1992, The Bakken Formation—An
integrated geologic approach to horizontal drilling, in J. E. Schmoker, E. Coalson, and C. Brown, eds.,
Geologic studies relevant to horizontal drilling: Examples from western North America: Denver,
Colorado, Rocky Mountain Association of Geologists, p. 215–226.
Chen, Z., K. G. Osadetz, C. Jiang, and M. Li, 2009, Spatial variation of Bakken or Lodgepole oils in the
Canadian Williston Basin: AAPG Bulletin, v. 93, no. 6, p. 829–851.
Cobb, D. G., 2013, Characterization of thickness anomalies in the Bakken and Three Forks formations,
north central North Dakota, USA: MS Thesis, Colorado School of Mines, 145 p.
Cuomo, M. C., and P. R. Bartholomew, 1991, Pelletal black shale fabrics: their origin and significance:
Geological Society, London, Special Publications, v. 58, no. 1, p. 221–232.
De Wever, P., and F. Baudin, 1996, Palaeogeography of radiolarite and organic-rich deposits in Mesozoic
Tethys: Geologische Rundschau, v. 85, no. 2, p. 310–326.
De Wever P, P. Dumitrica, J. P., Caulet, C. Nigrini C, M. Caridroit, 2002, Radiolarians in the sedimentary
record. Amsterdam: Gordon Breach Science Publishers, p. 533.
Egenhoff, S. O., and N. S. Fishman, 2013, Traces in the dark-sedimentary processes and facies gradients in
the upper shale member of the Upper Devonian-Lower Mississippian Bakken Formation, Williston
Basin, North Dakota, U.S.A.: Journal of Sedimentary Research, v. 83, no. 9, p. 803–824.
Fisher, L., M. F. Gazley, A. Baensch, S. J. Barnes, and G. Duclaux, 2014, Resolution of geochemical and
lithostratigraphic complexity: a workflow for application of portable X-ray fluorescence to mineral
exploration: Geochemistry: Exploration, Environment, Analysis, v. 14, no. 2, p 149-159.

110
Fishman, N. S., S. O. Egenhoff, A. R. Boehlke, and H. A. Lowers, 2015, Petrology and diagenetic history
of the upper shale member of the Late Devonian-Early Mississippian Bakken Formation, Williston
Basin, North Dakota: Geological Society of America Special Paper, v. 515, no. 7, p. 125–151.
Garrison, R. E., and A. G. Fischer, 1969, Deep water limestones and radiolarites of the Alpine Jurassic:
Depositional Environments in Carbonate Rocks: Special Publications of SEPM, v. 14, p. 193–198.
Gerhard, L. C., S. B. Anderson, and D. W. Fischer, 1990, Petroleum geology of the Williston Basin: Interior
cratonic basins: AAPG Memoir, v. 51, p. 507–559.
Gerhard, L. C., S. B. Anderson, J. A. LeFever, and C. G. Carlson, 1982, Geological Development, Origin,
and Energy Mineral Resources of Williston Basin, North Dakota: American Association of Petroleum
Geologist Bulletin, v. 66, no. 8, p. 989–1020.
Ghadeer, S. G., and J. H. S. Macquaker, 2011, Sediment transport processes in an ancient mud-dominated
succession: a comparison of processes operating in marine offshore settings and anoxic basinal
environments: Journal of the Geological Society, v. 168, no. 5, p. 1121–1132.
Ghadeer, S. G., and J. H. S. Macquaker, 2012, The role of event beds in the preservation of organic carbon
in fine-grained sediments: Analyses of the sedimentological processes operating during deposition of
the Whitby Mudstone Formation (Toarcian, Lower Jurassic) preserved in northeast England: Marine
and Petroleum Geology, v. 35, no. 1, p. 309–320.
Golonka, J., M. I. Ross, and C. R. Scotese, 1994, Phanerozoic Paleogeographic and Paleoclimatic Modeling
Maps: Canadian Society of Petroleum Geologists Memoir 17, v. 17, p. 1–47.
Hall, G., A. Buchar, and G. Bonham-carter, 2011, Quality control assessment of portable XRF analysers:
development of standard operating procedures, performance on variable media and recommended
uses. CAMIRO Project 10E01 report.
Haq, B. U., and S. R. Schutter, 2008, A chronology of Paleozoic sea-level changes.: Science (New York,
N.Y.), v. 322, no. October, p. 64–68.
Hartel, T. H. D., B. C. Richards, and C. W. Langenberg, 2012, Wabamun, Bakken equivalent Exshaw and
Banff formations in core, cuttings and outcrops from southern Alberta: CSPG-CSEG-CWLS
Convention: vision, p. 1–19.
Hatch, J. R., and J. S. Leventhal, 1992, Relationship between inferred redox potential of the depositional
environment and geochemistry of the Upper Pennsylvanian (Missourian) Stark Shale Member of the
Dennis Limestone, Wabaunsee County, Kansas, U.S.A.: Chemical Geology, v. 99, no. 1–3, p. 65–82.
Hayes, M. D., 1985, Conodonts of the Bakken Formation (Devonian and Missippian), Williston Basin,
North Dakota: The Mountain Geologist, v. 22, no. 2, p. 64–77.
Helz, G. R., C. V. Miller, J. M. Charnock, J. F. W. Mosselmans, R. A. D. Pattrick, C. D. Garner, and D. J.
Vaughan, 1996, Mechanism of molybdenum removal from the sea and its concentration in black
shales: EXAFS evidence: Geochimica et Cosmochimica Acta, v. 60, no. 19, p. 3631–3642.
Jarvie, D. M., R. J. Hill, T. E. Ruble, and R. M. Pollastro, 2007, Unconventional shale-gas systems: The
Mississippian Barnett Shale of north-central Texas as one model for thermogenic shale-gas
assessment: AAPG bulletin, v. 91, no. 4, p. 475–499.
Jin, H., 2014, Source rock potential of the Bakken shales in the Williston Basin, North Dakota and Montana:
PhD Thesis, Colorado School of Mines, 220 p.
Jin, H., and S. A. Sonnenberg, 2012, Source Rock Potential of the Bakken Shales in the Williston Basin,
North Dakota and Montana: Search and Discovery #20156, p. 0–4.

111
Johnson, R. L., 2013, The pronghorn member of the Bakken Formation, Williston Basin, USA: Lithology,
stratigraphy, reservoir properties: MS Thesis, Colorado School of Mines, 166 p.
Jones, B., and D. A. C. Manning, 1994, Comparison of geochemical indices used for the interpretation of
palaeoredox conditions in ancient mudstones: Chemical Geology, v. 111, no. 1–4, p. 111–129.
Jones, D. L., and B. Murchey, 1986, Geologic Significance of Paleozoic and Mesozoic Radiolarian Chert:
Annual Review of Earth and Planetary Sciences, v. 14, p. 455–492.
Karma, R., 1991, Conodonts of the Bakken Formation (Devonian-Mississippian) in Saskatchewan, northern
Williston Basin: Sixth International Williston Basin Symposium, v. 11, p. 70–73.
Kent, D. M., and J. E. Christopher, 1994, Geological history of the Williston Basin and Sweetgrass arch:
Geological atlas of the Western Canada Sedimentary Basin: Canadian Society of Petroleum Geologists
and Alberta Research Council, p. 421–429.
Kents, P., 1959, Three Forks and Bakken stratigraphy in west central Saskatchewan: Saskatchewan
Department of Mineral Resources, Report 37, 39 p.
Klinkhammer, G. P., and M. R. Palmer, 1991, Uranium in the oceans: Where it goes and why: Geochimica
et Cosmochimica Acta, v. 55, no. 7, p. 1799–1806.
Kocman, K. B., 2014, Interpreting Depositional and Diagenetic Trends in the Bakken Formation Based on
Handheld X-Ray Fluorescence Analysis, Mclean, Dunn, and Mountrail Countries, North Dakota: MS
thesis, Colorado School of Mines, 156 p.
Kohlruss, D. and Nickel, E., 2009, Facies analysis of the Upper Devonian-Lower Mississippian Bakken
Formation, southeastern Saskatchewan: in: Summary of investigations 2009, v. 1, Saskatchewan
Geological Survey, Saskatchewan Ministry of Energy and Resources, Misc. Rep. 2009-4.1, Paper A-
6, 11 p.
Kowalski, B., 2010, Quantitative mineralogic analysis of the middle Bakken member, Parshall Field,
Mountrail County, North Dakota: MS Thesis, Colorado School of Mines, 126 p.
Lash, G. G., and D. R. Blood, 2014, Organic matter accumulation, redox, and diagenetic history of the
Marcellus Formation, southwestern Pennsylvania, Appalachian basin: Marine and Petroleum
Geology, v. 57, p. 244–263.
LeFever, J. A, R. D. LeFever, and S. H. Nordeng, 2011, Revised Nomenclature for the Bakken Formation
(Mississippian-Devonian), North Dakota: in J. W. Robinson, J. A. LeFever, S. B. Gaswirth, eds., The
Bakken-Three Forks Petroleum System in the Williston Basin: Rocky Mountain Association of
Geologists, 2011, p. 11-26.
LeFever, J. A., C. D. Martinuik, E. F. R. Dancsok, and P. A. Mahnic, 1991, Petroleum Potential of the
Middle Member, Bakken Formation, Williston Basin: in J. E. Christopher and F. Haidl, eds.,
Proceedings of the Sixth International Williston Basin Symposium: Saskatchewan Geological Society,
Special Publication 11, p. 74-94.
Leonowicz, P., 2016, Nearshore transgressive black shale from the Middle Jurassic shallow-marine
succession from southern Poland: Facies, v. 62, no. 2, p. 1–23.
Lewan, M. D., and J. B. Maynard, 1982, Factors controlling enrichment of vanadium and nickel in the
bitumen of organic sedimentary rocks: Geochimica et Cosmochimica Acta, v. 46, no. 12, p. 2547–
2560.
Li, Y., and J. Schieber, 2015, On the origin of a phosphate enriched interval in the Chattanooga Shale
(Upper Devonian) of Tennessee-A combined sedimentologic, petrographic, and geochemical study:
Sedimentary Geology, v. 329, p. 40–61.

112
Lohr, S. C., and M. J. Kennedy, 2015, Micro-trace fossils reveal pervasive reworking of Pliocene sapropels
by low-oxygen-adapted benthic meiofauna.: Nature communications, v. 6, p. 6589.
Longman, M., K. Kocman, and L. Wray, 2014, Petrography of the Bakken black “shales” in the eastern
Williston Basin of North Dakota: AAPG Datapages/Search and Discovery Article #90193, Rocky
Mountain Section AAPG Annual Meeting, Denver, Colorado, July 20–22.
Loucks, R. G., and S. C. Ruppel, 2007, Mississippian Barnett Shale: Lithofacies and depositional setting of
a deep-water shale-gas succession in the Fort Worth Basin, Texas: AAPG Bulletin, v. 91, no. 4, p.
579–601.
Lyons, T. W., and R. A. Berner, 1992, Carbon-sulfur-iron systematics of the uppermost deep-water
sediments of the Black Sea: Chemical Geology, v. 99, no. 1, p. 1–27.
MacDonald, G. H., 1956, Subsurface Stratigraphy of the Mississippian Rocks of Saskatchewan: Geological
Survey of Canada Memoir 22, Queen’s Printer and Controller of Stationary, Ottawa 46 p.
Macquaker, J. H. S., and A. E. Adams, 2003, Maximizing Information from Fine-Grained Sedimentary
Rocks: An Inclusive Nomenclature for Mudstones: Journal of Sedimentary Research, v. 73, no. 5, p.
735–744.
MacQuaker, J. H. S., and R. L. Gawthorpe, 1993, Mudstone lithofacies in the Kimmeridge Clay Formation,
Wessex Basin, southern England; implications for the origin and controls of the distribution of
mudstones: Journal of Sedimentary Research, v. 63, no. 6, p. 1129–1143.
Macquaker, J. H. S., M. a Keller, and S. J. Davies, 2010, Algal Blooms and “Marine Snow”: Mechanisms
That Enhance Preservation of Organic Carbon in Ancient Fine-Grained Sediments: Journal of
Sedimentary Research, v. 80, no. 11, p. 934–942.
Macquaker, J. H. S., K. G. Taylor, and R. L. Gawthorpe, 2007, High-Resolution Facies Analyses of
Mudstones: Implications for Paleoenvironmental and Sequence Stratigraphic Interpretations of
Offshore Ancient Mud-Dominated Successions: Journal of Sedimentary Research, v. 77, no. 4, p. 324–
339.
Mainali, P., 2011, Chemostratigraphy and the paleoceanography of the Bossier-Haynesville Formation,
East Texas Basin, TX and LA, MS Thesis, University of Texas at Arlington, 90 p.
März, C., S. W. Poulton, B. Beckmann, K. Küster, T. Wagner, and S. Kasten, 2008, Redox sensitivity of P
cycling during marine black shale formation: Dynamics of sulfidic and anoxic, non-sulfidic bottom
waters: Geochimica et Cosmochimica Acta, v. 72, no. 15, p. 3703–3717.
McCabe, H. R., 1959, Mississippian Stratigraphy of Manitoba: Manitoba Department of Mines and Natural
Resources, Mines Branch Publication, v. 22, p. 58-1.
McLennan, S. M., 2001, Relationships between the trace element composition of sedimentary rocks and
upper continental crust: Geochemistry, Geophysics, Geosystems, v. 2, no. 4 (paper# 2000GC000109).
Middleton, N. J., and A. S. Goudie, 2001, Saharan dust: sources and trajectories: Transactions of the
Institute of British Geographers, v. 26, no. 2, p. 165–181.
Morford, J. L., W. R. Martin, R. François, and C. M. Carney, 2009, A model for uranium, rhenium, and
molybdenum diagenesis in marine sediments based on results from coastal locations: Geochimica et
Cosmochimica Acta, v. 73, no. 10, p. 2938–2960.
Nakamura, K., 2015, Chemostratigraphy of the Late Cretaceous Western Interior (Greenhorn, Carlile, and
Niobrara Formations), Denver Basin, CO, U.S.A: PhD Thesis, Colorado School of Mines, 129 p.

113
Nance, H. S., and H. D. Rowe, 2015, Eustatic controls on stratigraphy, chemostratigraphy, and water mass
evolution preserved in a Lower Permian mudrock succession, Delaware Basin, west Texas, USA:
Interpretation, v. 3, no. No. 1 (February), p. SH11-SH25.
Neira, C., J. Sellanes, L. A. Levin, and W. E. Arntz, 2001, Meiofaunal distributions on the Peru margin:
relationship to oxygen and organic matter availability: Deep-Sea Research I, v. 48, p. 2453–2472.
Parrish, J. T., 1982, Upwelling and Petroleum Source Beds, with Reference to Paleozoic.: AAPG Bulletin,
v. 66, no. 6, p. 750–774.
Passey, Q. R., K. Bohacs, W. L. Esch, R. Klimentidis, and S. Sinha, 2010, From Oil-Prone Source Rock to
Gas-Producing Shale Reservoir - Geologic and Petrophysical Characterization of Unconventional
Shale Gas Reservoirs, in International Oil and Gas Conference and Exhibition in China: Society of
Petroleum Engineers, (paper #131350-MS).
Peters, S. E., and D. P. Loss, 2012, Storm and fair-weather wave base: A relevant distinction? Geology, v.
40, no. 6, p. 511–514.
Raiswell, R., F. Buckley, R. A. Berner, and T. F. Anderson, 1988, Degree of pyritization of iron as a
paleoenvironmental indicator of bottom-water oxygenation: Journal of Sedimentary Research, v. 58,
p. 812–819.
Ramkumar, M., 2015, Toward Standardization of Terminologies and Recognition of Chemostratigraphy as
a Formal Stratigraphic Method: in M. Ramkumar, ed., Chemostratigraphy Concepts, Techniques, and
Applications, Elsevier, p. 1–21.
Richards, B. C., 1989, Upper Kaskaskia sequence: uppermost Devonian and lower Carboniferous, Chapter
9, in B. D. Ricketts, ed., Western Canada Sedimentary Basin, A Case History, Canadian Society of
Petroleum Geologists, Calgary, p. 165 – 201.
Rimmer, S. M., J. A. Thompson, S. A. Goodnight, and T. L. Robl, 2004, Multiple controls on the
preservation of organic matter in Devonian-Mississippian marine black shales: Geochemical and
petrographic evidence: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 215, no. 1–2, p. 125–
154.
Ross, D. J. K., and R. M. Bustin, 2009, Investigating the use of sedimentary geochemical proxies for
paleoenvironment interpretation of thermally mature organic-rich strata: Examples from the
Devonian-Mississippian shales, Western Canadian Sedimentary Basin: Chemical Geology, v. 260, no.
1–2, p. 1–19.
Sandberg, C. A., F. G. Poole, and J. G. Johnson, 1988, Upper Devonian of western United States: Canadian
Society of Petroleum Geologists, Memoir 14, p. 183–220.
Sano, J. L., K. T. Ratcliffe, and D. R. Spain, 2013, Chemostratigraphy of the Haynesville Shale: in U.
Hammes, A. Gale, eds., In Geology of the Haynesville Gas Shale in East Texas and West Louisiana,
AAPG Memoir 105, p. 137–154.
Schieber, J., 2009, Discovery of agglutinated benthic foraminifera in Devonian black shales and their
relevance for the redox state of ancient seas: Palaeogeography, Palaeoclimatology, Palaeoecology, v.
271, no. 3–4, p. 292–300.
Schieber, J., 2016, Mud re-distribution in epicontinental basins - Exploring likely processes: Marine and
Petroleum Geology, v. 71, p. 119–133.
Schieber, J., 2011, Reverse engineering mother nature? Shale sedimentology from an experimental
perspective: Sedimentary Geology, v. 238, no. 1–2, p. 1–22.

114
Schieber, J., 2014, Traces in the Dark — Sedimentary Processes and Facies Gradients in the Upper
Devonian – Lower Mississippian Upper Shale Member of the Bakken Formation, Williston Basin,
North Dakota, U.S.A.—Discussion: Journal of Sedimentary Research, v. 84, p. 837–838.
Schieber, J., D. Krinsley, and L. R. Riciputi, 2000, Diagenetic origin of quartz silt in mudstones and
implications for silica cycling.: Nature, v. 406, no. August, p. 981–985.
Schieber, J., J. B. Southard, and K. G. Thaisen, 2007, Accretion of mudstone beds from migrating floccule
ripples: Science, New Series, v. 318, no. 5857, p. 1760–1763.
Scott, C., and T. W. Lyons, 2012, Contrasting molybdenum cycling and isotopic properties in euxinic
versus non-euxinic sediments and sedimentary rocks: Refining the paleoproxies: Chemical Geology,
v. 324–325, p. 19–27.
Scott, C., J. F. Slack, and K. D. Kelley, 2017, The hyper-enrichment of V and Zn in black shales of the Late
Devonian-Early Mississippian Bakken Formation (USA): Chemical Geology, v. 452, p. 24–33.
Shanks, A. L., 2002, The abundance, vertical flux, and still-water and apparent sinking rates of marine snow
in a shallow coastal water column: Continental Shelf Research, v. 22, no. 14, p. 2045–2064.
Shaw, A.B., 1964. Time in Stratigraphy. McGraw-Hill, New York, N.Y., 365 pp.
Smith, M. G., and R. M. Bustin, 2000, Late Devonian and Early Mississippian Bakken and Exshaw Black
Shale Source Rocks, Western Canada Sedimentary Basin: A Sequence Stratigraphic Interpretation:
AAPG Bulletin, v. 84, no. 7, p. 940–960.
Smith, M. G., and R. M. Bustin, 1996, Lithofacies and paleoenvironments of the Late Devonian and Early
Mississippian Bakken Formation, Williston Basin: Bulletin of Canadian Petroleum Geology, v. 44,
no. 3, p. 495–507.
Smith, M. G., and R. M. Bustin, 1998, Production and preservation of organic matter during deposition of
the Bakken Formation (Late Devonian and Early Mississippian), Williston Basin: Palaeogeography,
Palaeoclimatology, Palaeoecology, v. 142, no. 3–4, p. 185–200.
Smith, M. G., R. M. Bustin, and M. L. Caplan, 1995, Sequence Stratigraphy of the Bakken and Exshaw
Formations: A Continuum of Black Shale Formations in the Western Canada Sedimentary Basin:
Seventh International Williston Basin Symposium, p. 399–409.
Sonnenberg, S. A., 2015, The Bakken Formation of the Williston Basin: A North American Success Story,
Unpublished.
Sonnenberg, S., 2014, The Upper Bakken Shale Resource Play, Williston Basin: Proceedings of the 2nd
Unconventional Resources Technology Conference, v. 10625, no. August, p. 25–27.
Sonnenberg, S. A., and A. Pramudito, 2009, Petroleum geology of the giant Elm Coulee field, Williston
Basin: AAPG Bulletin, v. 93, no. 9, p. 1127–1153.
Sperr, J. T., 1990, Exploration Models for Bakken Reservoirs: Williston Basin, North Dakota and Montana:
Montana Geological Society Bakken Workshop, p. 143–149.
Stasiuk, L. D., and M. G. Fowler, 2004, Organic facies in Devonian and Mississippian strata of Western
Canada Sedimentary Basin: Relation to kerogen type, paleoenvironment, and paleogeography:
Bulletin of Canadian Petroleum Geology, v. 52, no. 3, p. 234–255.
Theloy, C., 2014, Integration of Geological and Technological Factors Influencing Production in the
Bakken Play, Williston Basin: PhD Thesis Colorado School of Mines, 241 p.
Thermo Scientific NITON ® XL3t NITON XL3t User Manual, 2002.

115
Tinnin, B. M., S. T. R. Darmaoen, 2015, Chemostratigraphic variability of the Eagle Ford Shale, south
Texas: Insights into paleoredox and sedimentary facies changes, in J. A. Breyer, ed., The Eagle Ford
Shale: A renaissance in U.S. oil production: AAPG Memoir 110, p. 259–283.
Tribovillard, N., T. J. Algeo, T. Lyons, and A. Riboulleau, 2006, Trace metals as paleoredox and
paleoproductivity proxies: An update: Chemical Geology, v. 232, no. 1–2, p. 12–32.
Tribovillard, N., A. Riboulleau, T. Lyons, and F. Baudin, 2004, Enhanced trapping of molybdenum by
sulfurized marine organic matter of marine origin in Mesozoic limestones and shales: Chemical
Geology, v. 213, no. 4, p. 385–401.
Turner, J. T., 2015, Zooplankton fecal pellets, marine snow, phytodetritus and the ocean’s biological pump:
Progress in Oceanography, v. 130, p. 205–248.
Turner, J. T., 2002, Zooplankton fecal pellets, marine snow and sinking phytoplankton blooms.: Aquatic
Microbial Ecology, v. 27, p. 57–102.
Turner, B. W., J. A. Tréanton, and R. M. Slatt, 2016, The use of chemostratigraphy to refine ambiguous
sequence stratigraphic correlations in marine mudrocks. An example from the Woodford Shale,
Oklahoma, USA: Journal of the Geological Society, v. 173, no. 5, p. 854–868.
Turner, J. N, A. F. Jones, P. A. Brewer, M. G. Macklin, and S. M. Rassner, 2015, Micro-XRF applications
in fluvial sedimentary environments of Britain and Ireland: progress and prospects, in I.W. Croudace,
R.G. Rothwell, eds., Micro-XRF Studies of Sediment Cores, Developments in Paleoenvironmental
Research 17, Springer, Netherlands, p. 227–265.
Vecsei, A., W. Frisch, M. Pirzer, and A. Wetzel, 1989, Origin and tectonic significance of radiolarian chert
in the Austroalpine rifted continental margin: Siliceous deposits of the Tethys and Pacific regions.
Springer, Berlin, p. 65–80.
Wang, Y. J., H. Luo, and J. C. Aitchison, 2006, Influence of the Frasnian-Famennian event on radiolarian
faunas: Eclogae Geologicae Helvetiae, v. 99, p. 127–132.
Webster, R. L., 1984, Petroleum source rocks and stratigraphy of the Bakken Formation in North Dakota:
in Woodward J., F. F. Meissner, J. L. Clayton, eds., Hydrocarbon source rocks of the greater Rocky
Mountain region, Rocky Mountain Association of Geologists, p. 57 -82.
Wedepohl, B. K. H., 1971, Environmental Influences on the Chemical Composition of Shales and Clays:
Physics and Chemistry of the Earth, v. 8, p. 307–333.
Wedepohl, K. H., 2004, The Composition of Earth's Upper Crust, Natural Cycles of Elements, Natural
Resources, in Elements and Their Compounds in the Environment: E. Merian, M. Anke, M. Ihnat, M.
Stoeppler, eds., Occurrence, Analysis and Biological Relevance, Second Edition, Wiley-VCH Verlag
GmbH, Weinheim, Germany.
Wignall, P. B., 1991, Model for transgressive black shales: Geology, v. 19, no. 2, p. 167–170.
Wright, G. N., M. E. McMechan, and D. E. G. Potter, 1994, Structure and architecture of the Western
Canada sedimentary basin: Geological atlas of the Western Canada sedimentary basin, v. 4, p. 25–40.
Zheng, Y., R. F. Anderson, A. Van Geen, and M. Q. Fleisher, 2002, Preservation of particulate non-
lithogenic uranium in marine sediments: Geochimica et Cosmochimica Acta, v. 66, no. 17, p. 3085–
3092.
Zheng, Y., R. F. Anderson, A. Van Geen, and J. Kuwabara, 2000, Authigenic molybdenum formation in
marine sediments: A link to pore water sulfide in the Santa Barbara Basin: Geochimica et
Cosmochimica Acta, v. 64, no. 24, p. 4165–4178.

116
 4. CHAPTER 4

FACTORS CONTROLLING ORGANIC RICHNESS OF UPPER BAKKEN SHALE,

WILLISTON BASIN

Organic richness in shale is primarily controlled by the interplay of three factors: organic
productivity, preservation, and dilution by sediments. This study is focused on understanding the impact of
these factors on controlling the organic richness of Upper Bakken Shale (UBS), for which the Total Organic
Carbon (TOC) varies between 3-20 wt.%. A conceptual model for the accumulation and preservation of
organic matter (OM) in UBS is also proposed. The framework of this conceptual model includes the
depositional setting of the basin, including basin morphology, paleogeography, paleoclimate, and probable
nutrient sources.

This study is based on the results of the following three analyses: 1) source rock analysis (SRA)
for TOC and OM characterization; 2) induced coupled plasma mass spectrometry (ICP-MS) for elemental
concentration measurements; and 3) stable isotope measurements for carbon and nitrogen. These analyses
were performed on 37 core chips of UBS from well Robert Trust 1-13H, which is located at the southern
depocenter of the Williston Basin. Elemental concentration of Si, Al, K, Ti, and Ca were used for proxying
the effect of dilution and preservation of OM by detrital and biogenic sediments. The concentration of the
redox-sensitive trace elements (MO, U) and degree of pyritization (DOP) were used as proxies for
understanding the variation in redox condition during the deposition of the UBS. Stable isotopes of organic
carbon and nitrogen were used to understand the OM type in the UBS and the paleoproductivity during its
deposition.

The results of this study suggest that TOC and the factors which controlled the organic content
varied in the southern depocenter of the basin during the deposition of the UBS. It is proposed that the basin
was semi-restricted, with a stratified water column, and regeneration of the biolimiting nutrients such as P
and N were the major source of nutrients. These conditions varied during the deposition of the UBS,
primarily due to a relative change in sea level, which resulted in four distinct sub-units in the UBS: 1a, 1b,
2a, and 2b. Influx of clay helped in quick preservation of OM an thereby had a positive effect on organic-
richness during the deposition of all four sub-units. Biogenic silica had a dilution effect during the
deposition of organic matter in sub-units 1a and 1b. Detrital dolomite and siliciclastic silt also had a dilution
effect during the deposition of sub-units 2a and 2b. Sub-units 1a and 1b were deposited in a strongly euxinic
condition, which existed in the bottom water and extended to the photic zone intermittently. This prevented
the degradation and resulted in better preservation of OM. During the deposition of sub-units 2a and 2b,

117
the conditions were predominantly sub-oxic to anoxic along the sediment-water interface, which resulted
in more degradation of OM. Paleoproductivity was high during the deposition of sub-units 1a and 1b, and
paleoproductivity declined during the deposition of sub-units 2a and 2b.

4.1 Introduction

Oil and gas production from shale plays has witnessed enormous growth over the last decade;
however, there are challenges in terms of developing a better understanding of shale geology, which would
assure the continuation of this growth. The organic richness and their controlling factors in both modern
and ancient marine sediments is a fundamental but poorly understood area of shale geology. Currently,
there are two very different schools of thought which seek to explain the higher organic matter (OM)
accumulation in these organic-rich shales. The first theory emphasizes the role of increased primary
productivity of OM (Demaison and Moore, 1980) to explain the organic richness of these shales. The second
theory suggests that the primary reason for the organic richness in shales is the enhanced OM preservation
due to a strongly reducing paleoredox condition (Calvert, 1987; Pedersen and Calvert, 1990; Tyson, 2005).
However, recent studies have demonstrated that the total organic carbon (TOC) deposited in marine
sediments is actually a function of three variables: OM production, OM preservation and OM dilution by
detrital/biogenic sediments (Murphy et al., 2000; Werne et al., 2002; Sageman et al., 2003; Rimmer et al.,
2004; Bohacs et al., 2005; Tyson, 2005). These three variables are in turn controlled by a complex interplay
of multiple paleoceanographic and sedimentological factors, such as nutrient supply, sunlight,
sedimentation rate, benthic-redox condition, size and depth of the basin, the degree of basin restriction,
watermass mixing and changes in eustatic sea-level (Sageman et al., 2003; Bohacs et al., 2005). My study
is focused on understanding the role of these factors in controlling the organic richness of the UBS by
integrating the petrographic observations with the organic and inorganic geochemistry results.

The Early Mississippian UBS of the Williston Basin is a world-class source rock with an average
TOC of 11.3 wt.% in the U.S. portion of the basin (Webster, 1984; Smith and Bustin, 2000). The UBS is
an extensively studied shale system; the primary aims of these previous research efforts were to understand
the source rock potential, organic and inorganic geochemistry, and depositional environment of the UBS
(e.g. Meissner, 1978; Price et al., 1984; Webster, 1984; Hayes, 1985; Smith and Bustin, 1996, 1998;
Egenhoff and Fishman, 2013; Jin, 2014; Kocman, 2014; Longman et al., 2014). Based on the geochemical
analysis of the UBS, Smith, and Bustin (1998) concluded that enhanced organic productivity is the primary
reason for the organic richness of the UBS. The authors also proposed that during UBS deposition, organic
productivity varied, which in turn resulted in the observed temporal and spatial variation in OM richness of
the UBS. Smith and Bustin (1998) also suggested that the UBS was deposited in an anoxic stagnant basin,
which aided the preservation of OM. This interpretation was based on the presence of planar lamination,

118
abundant pyrite content and the lack of bioturbation and other body fossils. An anoxic stagnant condition
in the basin during the UBS deposition was also suggested by Price et al. (1984), Webster (1984), Hayes
(1985), Smith and Bustin (1996), Sonnenberg and Pramudito (2009), and Longman et al. (2014).

Smith and Bustin (1998) also used stable isotope of Nitrogen (N) as a proxy to determine
paleoproductivity and suggested that OM productivity was higher at the onset of the UBS deposition
compared to its levels in later stages. Moreover, based on the observation of spatial variation in TOC in the
UBS (as high as 35 wt.% in the Regina-Melville Platform area), the authors also proposed that OM
productivity varied regionally during the deposition. Smith (1996) used a quartz-to-illite and alkali feldspar-
to-illite ratio as a proxy to understand the change in detrital sedimentation rate in the UBS. The author
compared these ratios with TOC to understand the effect of siliciclastic sedimentation and dilution on OM
preservation; they observed no correlation between TOC vs. change in either the quartz-to-illite ratios or
TOC vs. alkali feldspar-to-illite ratios. Therefore, it was concluded that detrital sedimentation rate had
almost remained constant over the period of UBS deposition, and it did not affect the preservation/dilution
of TOC.

Scott et al. (2017), based on their inorganic geochemical analysis results, proposed that the
deposition of the UBS resulted in a rapid development of a strongly euxinic condition with very high
concentrations of dissolved H2S (~10mM). They further suggested that this euxinic condition mostly
remained close to the sediment-water interface and reached the photic zone episodically. These authors
referred to UBS samples having Zn concentrations of more than 500 ppm and TOC >7 wt.%, as “hyper-
enriched” with respect to Zn. It was proposed that hyper-enrichment of Zn reflects photic-zone euxinia and
the presence of photoautotrophic sulfide-oxidizing bacteria (PSOB). Based on this hypothesis, Scott et al.
(2017) concluded that UBS sections with hyper-enriched Zn concentration indicate the occasional
development of photic zone euxinia during their deposition. They also proposed that neither high primary
productivity nor a euxinic condition was the reason for the organic matter richness in the UBS. The authors
also suggested that the shallow depth of the basin (100-150m) resulted in a short transit time of the OM
through the water column, which after deposition degraded and resulted in the development of
anoxic/euxinic conditions.

Previous studies have reported the presence of the biomarker isorenieratene in Bakken shales
(Requeffo et al., 1991; Jiang et al., 2001). Isorenieratene is produced by PSOB Chlorobiaceae, which
requires a water column with light and a sulfidic condition to survive. Based on the presence of the
biomarker isorenieratene in Upper and Lower Bakken Shale, Requeffo et al., (1991) and Jiang et al., (2001)
suggested that euxinic conditions were extending up into the photic zone during deposition. On the other
hand, Egenhoff and Fishman (2013) proposed that the UBS was predominantly deposited in an overall

119
dysoxic condition (2.0-0.2ml O2/ l H2O according to Tyson and Pearson, 1991); their work was based on
petrographic studies, which indicated the presence of burrows and sedimentary structures, indicating
bedload transport processes. However, this study did not address the reservations regarding the reduced
preservation potential of OM during UBS deposition in a dysoxic condition.

Jin (2014) calculated the original TOC of UBS from SRA data and prepared the original TOC map
of the UBS. It was observed that along the eastern margin of the basin, the UBS is relatively lean in OM,
with lower than 10 wt.% TOC, while in the central and intermediate parts of the basin in the North Dakota
and eastern Montana areas, the UBS is generally organic-rich, with 17~20 wt.% original TOC. The author
suggested that this spatial variation in TOC content is probably due to the low OM production, dilution of
OM by terrigenous sediments, and poor preservation in a relatively dysoxic environment.

There is a clear disagreement over the interpretation of the redox conditions prevailing in the
Williston Basin during the UBS deposition. In the current study, I have integrated the petrographic
observations with organic and inorganic geochemistry results, which helped in providing a coherent
understanding of the paleoredox condition of UBS. Moreover, my research findings are not different from
the interpretation of Smith and Bustin (1998) which claimed constant anoxic conditions during the UBS
deposition and suggests that the redox condition of the basin was not constant and has varied from euxinic
to suboxic. The objectives of this study were as follows: 1) to develop an understanding of the interplay of
paleoproductivity, redox condition and dilution of OM in controlling the organic richness of UBS; 2) based
on this understanding, to provide a conceptual model for the UBS deposition. My results show that the
redox condition, which controls the preservation potential of OM, should be considered along with
paleoproductivity in order to understand the organic richness in the UBS. I also examined the effect on the
organic richness of OM dilution by detrital and biogenic sediments.

4.2 Methodology: Sampling and analytical procedures

The study area and geological background are same as those used in Chapter 3 and were discussed
in Section 3.2. The current study is based on the petrographic and geochemical analyses of the drill core of
well Robert Trust 1-13H (RT), which is located near the southern part of the basin depocenter (Figure 4.1).
In Well RT, the UBS thickness of 21.2ft is close to its maximum thickness in the basin; therefore, the core
from Well RT is likely to have captured almost the full depositional history of UBS in the southern part of
the basin depocenter. The core of well RT is curated at the North Dakota Geological Survey core repository
center at Grand Forks (ND). About 40 core chips were collected at every 0.5 to a 1ft interval of the core for
geochemical analysis, and thirteen paired thin-section billets were collected for petrographic thin section
preparation. The methodology for thin section preparation was the same as that presented in Section 3.3.4.

120
The geochemical analyses I used were the following: 1) source rock analysis (SRA) for TOC measurements;
2) inductively coupled plasma mass spectrometry (ICP-MS) for identification of major and trace element
concentrations; 3) Sulphur concentration measurements by an element analyzer; and 4) stable isotope
measurements for organic Carbon and Nitrogen. Prior to the geochemical analysis, the core chip samples
were cleaned with deionized water to remove salt crust and/or drilling mud residues, and subsequently, the
samples were air dried for 24 hours. The cleaned samples were crushed by mortar and pestle to a 200-mesh
size powder. Aliquots of thirty-seven samples were analyzed for TOC using the Weatherford Source Rock
Analyzer (SRA) instrument at Colorado School of Mines using the procedure described by Jin (2014).

Figure 4.1: Location of well Robert Trust 1-13H (RT) is shown by the red star in the isopach map of the
Upper Bakken Shale (modified from Sonnenberg et al., 2017).

Remaining aliquots of the powdered samples were used for determining major, minor and trace
element concentrations by the SGS Mineral Services, Canada. This involved performing ICP-AES and ICP-
MS analyses after sodium peroxide fusion of the powdered aliquots (SGS method GE ICM90A). ICP-AES
and ICP-MS analyses were conducted on a total of thirty-six samples. Details about the protocol used in
these analyses can be accessed at http://www.sgs.com/en/mining/analytical-services/geochemistry/
digestion-and-fusion. Sulphur concentration measurements were performed using the Costech ECS4010

121
elemental analyzer Colorado Plateau Stable Isotope Lab at Northern Arizona University. Carbon and
nitrogen isotope analyses were performed on the samples by the Stable Isotope Ratio Facility for
Environmental Research at the University of Utah. For this analysis, the powdered samples were treated
with 1mol/L hydrochloric acid (HCl) to remove the carbonate component. Thereafter, the samples were
rinsed five times with distilled water followed by centrifugation, which was followed by drying the samples
in an oven at a temperature of 50ºC. Finally, the dual C and N isotope (δ13Corg, δ15Norg) analyses were
performed on these samples using the DELTA plus Advantage isotope ratio mass spectrometer configured
through a CONFLO III interface for automated continuous-flow analysis. All values of δ13Corg and δ15Norg
are reported in per mille ‰ relative to V-PDB and air, respectively. The measurement precision was
assessed by a replicate of samples and internal lab standards and was determined to be ±0.15‰ for δ13C
and ±0.2‰ for δ15N.

The role of detrital and biogenic sediments influx in the preservation of OM was analyzed using
the geochemical proxies for detrital and biogenic sediments. The theoretical background for the
geochemical proxying for detrital and biogenic sediments was discussed in Section 3.3.3. In this study, the
stable isotope of Nitrogen (δ15Norg) was used as a proxy for paleoproductivity, whereas the stable isotope
of Carbon (δ14Corg) was used as a proxy for the OM type (terrestrial vs. marine). The paleoredox condition
was interpreted on the basis of different geochemical proxies, which are as follows: i) the absolute
concentration of trace elements (TE) Mo; ii) the covariation pattern of TOC with Mo and U; iii) the slope
of regression lines of TOC vs. the Mo crossplot; iv) the covariation pattern of the enrichment factor (EF)
of U and the EF of Mo; and v) the degree of pyritization calculated from C-S-Fe (Carbon-Sulphur-Iron)
systematics. These proxies have been used by various researchers for understanding the prevalent redox
conditions in modern and ancient marine systems (Sageman et al., 2003; Algeo and Maynard, 2004;
Rimmer et al., 2004; Tribovillard et al., 2006, 2012; Lash and Blood, 2014; Chen and Sharma, 2016). The
EF of an element in each sample is calculated by comparing the Al-normalized trace element ratio to that
of an average shale. In other words, the EF for element X (XEF) in a given sample is equivalent to the ratio
(X/Al)sample/(X/Al)AS, where subscript AS refers to average shale (Wedepohl, 1971; Tribovillard et al.,
2006). My study uses the nomenclature proposed by Tyson and Pearson (1991) to describe the redox
conditions. According to this nomenclature, the oxic condition refers to 8.0-2.0 ml/1 O2 in water, the
dysoxic condition refers to 2.0-0.2 ml/l O2 in water, the suboxic condition refers to 0.2-0.0 ml/1O2 in water,
and the anoxic condition refers to 0.0 ml/l O2 and 0 ml/l H2S in water. Euxinic or anoxic-sulfidic water
lacks O2 and contains free H2S (Raiswell and Berner, 1985). According to Tyson and Pearson (1991), the
term “bottom water” refers to the water column, which is less than 1 meter above the sediment-water
interface.

122
 Proxies for paleoredox condition

Trace Element covariation

The theoretical background for the use of trace elements U and Mo for interpreting the paleoredox
condition were discussed in Section 3.3.3. The trace metals Mo and U exhibit variability in an oxidation
state, solubility, and/or change in host-phase as a function of the redox condition of the depositional
environment. These redox-sensitive trace elements are more soluble and display a conservative behavior in
oxic conditions but are less soluble in reducing conditions, which results in their authigenic enrichments in
oxygen-depleted sedimentary facies. This makes these elements ideal proxies for evaluating redox
conditions in paleomarine systems (Algeo and Maynard, 2004; Tribovillard et al., 2006). The concentration
of these redox-proxying trace elements is very low in the continental crust, and they maintain a uniform
concentration of seawater globally because of their long residence time. These trace elements are also
present in low concentrations in plankton; therefore, their enrichment in marine sediments can be attributed
to their authigenic uptake from seawater due to changes in paleoredox conditions. Furthermore, these
elements are relatively immobile; therefore, their concentrations are generally not affected by diagenesis
and thermal maturation (Algeo and Rowe, 2012).

In oxic conditions, the trace-metal Mo in seawater is present in the form of stable and unreactive
molybdate ions (MoO − ). Marine sediments accumulating in oxic environment have limited authigenic
enrichment of Mo, and they exhibit typical detrital background values of Mo concentration, which range
from 1-5 ppm (Zheng et al., 2000; McLennan, 2001; Morford et al., 2009). However, in reducing
environments, in the presence of even very small amounts of dissolved HS − , MoO −
ions become reactive
and get transformed into a series of particle-reactive thiomolybdate ions (MoO� S −x ) (Helz et al., 1996;
Zheng et al., 2000). Mo in this form is rapidly sequestered either by getting adsorbed in sulfurized organic
matter or by forming a solid solution with Fe-S to get precipitated as authigenic sulfide minerals (Helz et
al., 1996; Tribovillard et al., 2004). At higher levels of dissolved hydrogen sulfide in an euxinic condition,
authigenic uptake and burial rate of Moauth increases by two to three times (Scott and Lyons, 2012). Most
of the Moauth uptake occurs either along the sediment/water interface or within the pore water of the
sediments (Zheng et al., 2000; Morford et al., 2009). Generally, low to moderate authigenic enrichment of
Mo and its correlation with TOC indicates an anoxic environment of deposition, while very high enrichment
of Mo indicates euxinic conditions with dissolved sulfide in water. However, along with the intensity of
reducing conditions, authigenic Mo enrichment in sediment is also dependent on the aqueous concentration
of Mo (source-ion availability) and the concentration of sedimentary organic matter (host-phase
availability) (Algeo and Lyons, 2006; Algeo and Rowe, 2012). This relationship can be expressed as
follows:

123
 [Mo]sed ≡ [TOC]sed * [Mo]aq (4.1)

[Mo]aq ≡ [Mo/TOC]sed, (4.2)

where the subscripts sed and aq refer to the sediment and aqueous concentration of Mo,
respectively. Based on this observation, Algeo and Lyons (2006) developed a method to understand the
degree of basin restriction. Eq. (2) indicates that the ratio of Mo to TOC in the sediments is proportional to
the concentration of aqueous Mo. Moreover, the concentration of aqueous Mo is dependent on the extent
of deepwater renewal in the basin, which in turn is controlled by the degree of basin restriction and basin
size. In restricted and silled basins with limited deepwater renewal, the uptake of Mo by sediment in a
reducing environment decreases the aqueous concentration of Mo slowly. This condition results in a
decrease in the rate of Mo enrichment in sediment over time, although the rate of OM accumulation may
remain high due to the reducing condition, which thereby results in low [Mo/TOC] sed. However, in
unrestricted open-marine basins, the aqueous concentration of Mo remains high due to strong deepwater
renewal, and this results in high [Mo/TOC]sed. Thus, the positive covariation pattern between Mo and TOC
concentration, specifically the slope m of the regression line in the Mo-TOC crossplot, exhibits a systematic
relationship with the degree of basin restriction and deepwater renewal (Algeo and Lyons, 2006; Algeo and
Rowe, 2012). Algeo and Lyons (2006) used this concept in studying modern silled basins at the Black Sea,
Framvaren Fjord, Cariaco Basin and Saanich Inlet, and observed that an m value less than 10 is the
characteristic of strong hydrographic restriction (Black Sea), whereas m between 10-25 represents moderate
restriction (Framvaren Fjord and Cariaco Basin), and m greater than 25 represents weak restriction (Saanich
Inlet). This concept was applied to various shale formations to study the ancient silled basins of the
Devonian-Carboniferous time (Algeo et al., 2007). The degree of basin restriction affects the deepwater
renewal time in the basin. Deepwater renewal time is dependent on [Mo]aq, which in turn is dependent on
the authigenic uptake of Mo by sediment. Algeo and Lyons (2006) used this background and estimated the
deepwater renewal time from Mo flux for modern anoxic basins.

Trace-metal U is present mainly as a soluble U(VI) in oxic-suboxic conditions and is reduced to

less soluble U(IV) in oxygen-depleted conditions. U concentration in average shale is around 3.7 ppm
(Wedepohl, 1991). Reduction of U commences at the Fe (II)- Fe(III) redox boundary and the U enrichment
is controlled by microbially-mediated Fe redox reactions rather than by the presence of HS- (Zheng et al.,
2002). The uptake of a reduced state of U� ℎ by the sediment primarily occurs through the formation of
organic metal ligands, or precipitation of crystalline uraninite or a metastable precursor to uraninite
(Klinkhammer and Palmer, 1991; Zheng et al., 2002). Thus, in comparison to Moauth, Uauth enrichment in
marine sediment starts at relatively less reducing conditions and at shallower depths within the sediment
column (Morford et al., 2009).

124
 These differences in geochemical behavior between Mo and U result in different relative
enrichment of Mo and U for specific redox conditions. Algeo and Tribovillard (2009) applied this concept
in modern marine settings and identified three unique covariation patterns of the EFs of U-Mo, which can
be associated with different redox conditions and the processes specific to marine settings. The authors also
observed that along with the EF covariation patterns of U and Mo, the (Mo/U)auth ratio of the sediment,
when compared to the (Mo/U)aq molar ratio in seawater (SW) of ~7.5-7.9, has the potential to provide new
insights regarding the paleoceanography of organic-rich depositional systems. The first covariation pattern
is characteristic of sediments from an unrestricted marine setting in the eastern tropical Pacific Ocean; it
exhibits greater relative Uauth enrichment but low or no Moauth at low EFs (Mo/Uauth ~0.1-0.3×SW) and
indicates a suboxic condition. However, for progressively greater relative Moauth enrichment at high EFs
(Moauth:Uauth ratios >1×SW), the condition is increasingly more anoxic to euxinic in the water column. The
second pattern is distinctive of sediments from the semi-restricted setting of the Cariaco Basin, which
exhibits strong Moauth relative to Uauth enrichment at all EFs (Moauth:Uauth ratios >>1 × SW, 3-10 × SW).
This highlights the role of the Mn-oxyhydroxide particulate shuttle, which helps to transfer only the
molybdate ions (not U ions) from the water column to the sediment-water interface and thereby enhances
the relative enrichment of Moauth. The third covariation pattern is characteristic of sediments deposited
below the pycnocline in the restricted and euxinic setting of the modern Black Sea. These sediments exhibit
higher relative Moauth enrichment than Uauth for samples with lesser degrees of total enrichment (Moauth:Uauth
ratios ~1 × SW) and progressively higher enrichment of Uauth relative to Moauth as total enrichment increases;
this indicates the evolution of sub-pycnoclinal watermass chemistry due to sustained euxinia, which draws
down the availability of aqueous Mo (source-ion) concentration with time (Algeo and Lyons, 2006; Algeo
and Tribovillard, 2009). The validity of this U-Mo covariation pattern-based approach for analyzing the
paleoredox conditions and processes during deposition of organic-rich shales were discussed by Algeo and
Tribovillard (2009), Tribovillard et al. (2012), and Lash and Blood (2014).

C-S-Fe relationship

C-S-Fe systematics are used to differentiate between the varying redox conditions during the
deposition of organic-rich shales. Dean and Arthur (1989), Arthur and Sageman (1994) and Rimmer et al.
(2004) suggested that ternary Fe-S-C diagrams can be used to estimate Degree of Pyritization (DOP). A
typical Fe-S-C ternary diagram used for this purpose is shown in Figure 4.2. Normal marine oxygenated
samples are plotted along a line equivalent to a S/C ratio of 0.4 on this ternary diagram, whereas samples
in which all iron is reactive and associated with pyritic sulfur (DOP=1), are plotted along the line having
an S/Fe ratio of 1.15 (based on the stoichiometry of pyrite) that intercepts the Fe-S axis at 0.54. DOP lines
and fields representing the dysoxic/suboxic and the anoxic conditions are also shown.

125
 Carbon vs. Sulfur (Corg-Spyr) crossplots are also used to distinguish normal marine and anoxic-
euxinic environments (Rimmer et al., 2004). In Corg-Spyr crossplot, a strong positive correlation between
these two variables is observed for normal marine sediments, which had accumulated below oxygenated
water and the trend line has a zero intercept in the crossplot (Raiswell and Berner, 1985). This is because
pyrite formation in normal marine sediments is controlled by the availability of OM. However, for euxinic
sediments, pyrite formation is primarily controlled by the availability of reactive iron; therefore, the Corg-
Spyr crossplot shows a non-zero sulfur intercept for the trend line, or it may show no correlation (Raiswell
and Berner, 1985). Similarly, DOP, which is the ratio of pyritic iron to reactive iron (pyritic iron plus HCl
soluble iron), is considered to be a reliable indicator for depositional redox conditions, especially in
environments that are not limited by reactive Fe (Raiswell et al., 1988; Jones and Manning, 1994). DOP
values less than 0.42 indicate normal marine oxygenated conditions; DOP values between 0.42 and 0.75
indicate dysoxic to anoxic conditions in a less stratified water column; and DOP values greater than 0.75
indicate anoxic conditions with strongly a stratified water column (Raiswell et al., 1988; Hatch and
Leventhal, 1992; Algeo and Maynard, 2004; Rimmer et al., 2004). Uniformly high DOP values are typically
interpreted to indicate deposition under euxinic conditions and suggest that nearly all of the available
reactive Fe was utilized in the formation of pyrite (Raiswell et al., 1988; Lyons and Berner, 1992).

Figure 4.2: Schematic S-Fe-C ternary diagram showing the different degrees of pyritization (DOP) regions
for suboxic, anoxic and euxinic conditions (modified from Rimmer et al., 2004).

126
 Proxy for Paleoproductivity and Organic Matter type

In our current study, δ15Norg was used as a proxy to assess the paleoproductivity in the UBS.
Nitrogen (N) in the form of nitrate is an essential biolimiting nutrient for marine productivity.
Phytoplanktons preferentially incorporate the lighter isotope of nitrogen (N14), which results in a relative
enrichment of the heavier isotope of nitrogen (N15) in water (Marchitto, 2007). Therefore, in regions where
the nitrate supply is heavily utilized, the surface ocean isotopic enrichment can be relatively large, which
results in heavier δ15Norg (Altabet and Francois, 1994). OM that has a light δ15Norg value reflects conditions
of high nutrient supply to the surface water relative to biological consumption, which indicates eutrophic
conditions. Heavy δ15N values reflect a more limited supply of nitrate relative to consumption, which is
indicative of oligotrophic conditions. Therefore, lighter δ15N values reflect an increase in the availability of
nutrients, which indicates high primary productivity at the time of deposition. Heavier δ15Norg values reflect
a decrease in availability of nutrients and thus indicate reduced primary productivity at the time of
deposition (Altabet and Francois, 1994; Altabet and Francois, 1994; Montoya, 1994). Caplan and Bustin
(1998) and Smith and Bustin (1998) used δ15Norg as a proxy to predict relative nutrient utilization and
paleoproductivity of the organic-rich shales from the Bakken and Exshaw Formations. A stable isotope of
organic carbon (δ13Corg) was used in their study as a proxy to determine changes in organic matter sources.
During the Devonian period, δ13Corg was found to be about −30.5‰ for marine OM and −25‰ to −26‰
for the terrestrial organic matter, indicating marine organic matter is lighter in δ13Corg (Maynard, 1981;
Jaminski et al., 1998). My study also used a modified Van Krevelen diagram (hydrogen index(HI) vs.
Tmax) to determine the OM type.

4.3 Results and Interpretation

This section presents the various experiments performed, their interpretation, and our results. I first
present the chemostratigraphy, core description and petrographic results, and elaborate these based on the
methods outlined in Section 3.3. I then discuss the results for TOC and OM types, obtained from SRA and
oxygen stable isotope experiments. The effect of dilution on the organic richness of the UBS based
elemental proxy is then described, and the results for various paleoredox proxying methods are interpreted.
Finally, the paleoproductivity during the UBS deposition of is interpreted from the nitrogen stable isotope
data.

Chemostratigraphy, Core description, and Petrography

Table 4.1 lists characteristic differences between the chemostratigraphic sub-units of the UBS in
terms of vertical distribution of lithofacies, chemostratigraphic profiles of both detrital proxying elements
(Si, Al, K, Ti, Ca, Zr/Rb) and redox-sensitive trace elements (Mo and U). Section 3.4.5 presented the

127
identification in well RT and 16 other wells of two chemostratigraphic UBS units: Unit-1 (lower) and Unit-
2 (upper), and two sub-units of Unit-1 namely, 1a and 1b. In this chapter, I present a close examination of
the trends in redox-sensitive trace elements for well RT, which helped me to further divide Unit-2 into sub-
units 2a and 2b. In this study, it was observed that Mo and U had limited enrichment in sub-units 2a, whereas
moderate to the high enrichment of these two redox-sensitive trace elements were observed in sub-unit 2b.
The vertical distribution of the different lithofacies is shown in the core description of well RT presented
in Figure 4.3. Here, the facies classification scheme and description methodology for the UBS is consistent
with those of Section 3.3.4. As discussed in Chapter-3, the UBS consists of six different facies; however,
in well RT, it primarily consists of only three facies: F1, F2, and F4. The characteristics and sedimentary
processes of these facies were discussed in Section 3.4.1.

Table 4.1: The characteristic differences between sub-units 1a, 1b, 2a and 2b in terms of the biogenic,
detrital, grain size proxy, redox proxy, and facies.
Sub- Biogenic/Detrital/ Redox proxy Facies
units Grainsize proxy
1a High Si and low Al, K, Highest Mo and U content Dominated by massive to very finely
1b and Ti, Ca and Zr/Rb High Mo and U, comparatively laminated silt-bearing mudstone (F2)
ratios compared to those less than that in 1a followed by siliceous mudstone (F1)
in Unit-2
2a Limited enrichment of Mo and Macrofossil-bearing silt-rich mudstone
Low Si and high Al, K,
U (F4)
and Ti Ca and Zr/Rb
2b Moderate enrichment of Mo Macrofossil-bearing silt-rich mudstone
ratios compared to those
and U (F4) and massive to very finely
in Unit-2
laminated silt-bearing mudstone (F2)

TOC, OM Characteristics, and Type

Table 4.2 lists the SRA-derived organic geochemistry results for four chemostratigraphic UBS sub-
units of the UBS found in Well RT. These results and the TOC profile is shown in Figure 4.3, which indicate
that TOC varies considerably among the different UBS sub-units in Well RT. For example, in sub-unit 1a,
average TOC is about 12.9 wt.%, while in sub-unit 2a, average TOC is about 5.0 wt.%. The effect of thermal
maturation of OM on measured TOC (TOC A) was corrected using the method described by Jarvie (2012),
and thereby original TOC (TOCo g ) was calculated. Most of our analyses which require TOC values were
performed using TOCo g . Jin (2014) suggested that the onset of hydrocarbon generation in Bakken shales is
marked by Tmax of 425ºC, while Tmax of 435ºC marks the peak of hydrocarbon generation, with Tmax of
445ºC being the last stage of the oil window. Therefore, as the UBS Tmax in Well RT is between 440º-
448ºC (average 445ºC), the location of well RT is in the oil maturity window.

128
Figure 4.3: Integrated profile of UBS in core Roberts Trust 1-13H (RT) showing different chemostratigraphic units of UBS, which were distinguished
based on the vertical distribution of lithofacies, detrital and biogenic proxying elements of Si, Al, K, Ti, Zr/Rb ratio, Ca and redox-proxying elements
of Mo and U. Characteristics of each chemostratigraphic sub-unit are summarized in Table 4.1. The profile of TOC, stable isotope of δ15Norg, which
is a proxy for the paleoproductivity, and stable isotope of δ13Corg, which acts as the proxy for the OM type, are also shown. TOC A represents TOC
measured from source rock analysis (SRA) and TOCo g represents original TOC calculated from TOC A. The basis for identifying the Maximum
flooding surface (MFS) was described in Section 3.5.5.

129
Table 4.2: Organic geochemistry results for different chemostratigraphic units of Well RT as derived from
the Source Rock Analysis (SRA). Legend: μ: mean; R: Range; σ: Standard deviation; HI: Hydrogen Index;
OI: Oxygen Index; TOC A: SRA measured TOC; TOCo g: Original TOC calculated TOC A.
Units TOC A wt. % HI mg HC⁄g OC OI mg HC⁄g OC TOCo g wt. %
R μ σ R μ σ R μ σ R μ σ
1a 10-17 12.9 1.9 248-350 312 26 1-3 2.0 0.6 13-20 16.2 2.2
1b 9-14 10.8 1.7 249-326 273 26 1-4 2.2 0.7 13-17 14.2 1.8
2a 3-9 5.0 2.1 129-249 184 41 5-10 7.6 2.5 3-11 5.5 2.7
2b 6-12 8.7 2.5 208-297 254 33 2-6 4.3 1.3 8-16 11.1 2.9

The TOC data for RT well were plotted on the modified Van Krevelen diagram shown in Figure
4.4. Both Table 4.2 and Figure 4.4 show that for sub-unit 2a, the average value of HI (184) is considerably
lower than those in other sub-units (HI: 312-254). Such variation in HI can be explained either by variation
in kerogen type (marine vs. terrestrial) or an alteration in thermal maturation (Tissot and Welte, 1984).
However, in Well RT, Tmax varies in a narrow range (440ºC to 448ºC), which cannot justify the observed
degree variation in HI. Therefore, the comparatively low average HI observed in sub-unit 2a suggests that
this sub-unit consists of a different kerogen type. Figure 4.4 also shows that sub-units 1a, 1b, and 2b
primarily consist of Type II-III OM; however, 2a has a greater amount of Type III (terrestrial) OM along
with Type II OM. Jin (2014) suggested that the UBS primarily consisted of Type II and II-III kerogens
near the basin depocenter, while inputs of type III OM are present exclusively along the eastern margins of
the basin. However, type III (terrestrial) OM has not been reported for the UBS from the basin depocenter.
For example, Stasiuk (1993), based on organic petrography, concluded that low maturity Bakken shales
from Canada mainly consist of: i) bituminite, comprised of dark brown fluorescent to non-fluorescent
amorphous granular groundmass, which forms micro-laminae, and ii) unicellular algae (alginite), which
showed textures indicating partially degraded bituminite. Therefore, the author suggested that the Bakken
shales primarily consisted of Type II kerogen. Hackley and Cardott (2016) studied matured UBS samples
from the southern edge of the UBS in the Billings Nose area and observed that the UBS primarily consisted
of non-fluorescent groundmass of solid bitumen with some fine-grained inertinite (wind-blown char).

Box and whisker plot in Figure 4.5 shows the result for the stable isotope analysis for δ13Corg for
the various chemostratigraphic sub-units of UBS in Well RT. Figure 4.3 shows the profile of δ13Corg in the
UBS interval, which demonstrates its variation among the different chemostratigraphic sub-units of the
UBS in Well RT. The average δ13Corg levels in sub-units 1a and 1b are -30.1‰ and -29.7‰; this is close to
the average δ13Corg of -30‰ observed in Type II OM (marine) in the Late Devonian and Early Mississippian.
This suggests that the OM in sub-units 1a and 1b are of marine origin. The average δ13Corg of sub-unit 2a is
-25.4‰, which suggests sub-unit 2a is dominated by Type III OM (terrestrial), as the terrestrial OM have
slightly heavier δ13Corg (-25‰ to -26 ‰). The average δ13Corg of sub-unit 2b is -28.1‰, which suggests that

130
sub-unit 2b has Type II-III OM. Therefore, the results of the stable isotope of C are consistent with the SRA
data and further confirm that sub-unit 1a and 1b consist of marine (Type II) OM, while the OM in sub-unit
2a is mostly of terrestrial origin.

Figure 4.4: Samples from Well RT plotted on the modified Van Krevelen diagram showing sub-unit 1a, 1b
and 2b predominantly consist of type II-III organic matter, while sub-unit 2a has a higher proportion of type
III organic matter.

Effect of Detrital and Biogenic Sediment Influx on OM Richness

Principal component analysis (PCA) presented in Section 3.4.2 suggested that Al, K, and Ti in the
UBS are of detrital origin; therefore, in our current study, the concentration of these elemental was used to
study the effect of dilution by detrital influx on OM richness in the UBS. Figure 4.6 shows Al and K have
strong positive covariance with Ti, which indicates that Al and K has a detrital origin in the UBS for both
Unit-1 and Unit-2. Moreover, both Al and K can be used as proxies for detrital illite clay content in the

131
UBS, as described in Section 3.4.1. As was shown in Figure 4.3, Unit-2 has a higher concentration of the
detrital proxying elements Al, K and Ti compared to those in Unit-1. However, the concentrations of Al, K
and Ti in sub-units 1a and 1b do not vary considerably; therefore, these sub-units were considered as one
single unit (Unit-1) for the purposes of analyzing the effect of detrital sediment influx on OM richness.
Likewise, sub-units 2a and 2b were considered as one single unit (Unit-2) for this analysis.

Figure 4.5: Box and whisker plot showing the stable isotope of carbon (δ13Corg) results for the different
chemostratigraphic UBS units in Well RT. Mean values of δ13Corg for each of the chemostratigraphic sub-
units are labeled to the right of each of the boxes. Average values for δ13Corg terrestrial-type III (-25‰ to -
26 ‰) and marine-type II (-30 ‰) OM are also shown from Chen and Sharma (2016) by the green rectangle
and brown line respectively.

As shown in crossplots between TOC with Al and K for Unit-1 in Figure 4.6 Al and K show no
covariance with TOC, which indicates that the influx of clay did not affect OM richness in Unit-1. As Al
and K do not have negative correlation in their respective crossplots with ������ , and according to Ibach
(1982) and Werne et al. (2002), this suggests that the influx of clay did not dilute OM. However, as shown
in Figure 4.6, ������ correlates positively with both Al and K for Unit-2, which, according to Rimmer et
al. (2004) and Passey et al. (2010), indicates that an influx of clay helped in preservation of organic matter
as it was buried quickly. Alternatively, a clay surface area may also have played a role in OM preservation
because this would provide a key site for OM adsorption and accumulation in organic-rich sediments
(Kennedy et al., 2002).

Figure 4.3 shows that Unit-2 has a higher concentration of Ca than that in Unit-1, which is
dominated by macrofossil-bearing silt-rich mudstone (F4). As discussed in Section 3.4.1, core description

132
and the petrographic study confirmed the presence of articulate and disarticulated shells and detrital
dolomites in facies F4. Therefore, Ca was used as a proxy to determine the effect of silt-sized detrital
dolomite and bioclasts in the OM accumulation. As shown in Figure 4.7, the crossplot between TOC and
Ca does not have any correlation for Unit-1, which suggests that the influx of carbonate sediments was
minimum and had no influence in OM accumulation. However, TOC and Ca have a negative correlation
for Unit-2, indicating an influx of silt-sized detrital dolomite and the presence of shell fragments, which
has a detrimental effect on the preservation of OM.

Figure 4.6: Cross plot between Ti and Al show a positive linear covariance for Unit-1, which indicates that
Al has a detrital origin. (b) Crossplot between Ti and K show a positive linear covariance for Unit-1, which
indicates that K has a detrital origin, and for this unit, K can be used as a proxy for detrital illite clay. (c)
Crossplot between Al and ������ show weak positive covarriance, which indicates that influx of clay did
not result in dilution of organic matter as Al is a proxy for detrital clay minerals in UBS (Section 3.4.2,
Chapter-3). (d) Crossplot between K and ������ show no covarricance, which confirms that influx of clay
did not result in dilution of organic matter as K is a proxy for detrital clay minerals in UBS (Section 3.4.2,
Chapter-3). (e) Crossplot between Ti and Al show a positive linear covariance for Unit-2, which indicates
that Al has a detrital origin in this unit also. (f) Crossplot between Ti and K show a positive linear covariance
for Unit-2, which indicates that K has a detrital origin, and for this unit, K can be used as a proxy for dertrital
illite clay. (g) Crossplot between Al and ������ for Unit-2 show a positive linear covariance, which
indicates that influx of clay helped in organic matter preservation in Unit-2 as Al is used as proxies for
detrital clay minerals (h) Crossplot between K and ������ for Unit-2 show a positive linear covariance,
which confirms that influx of clay helped in organic matter preservation in Unit-2.

133
Figure 4.7: (a) Crossplot between Ca and ������ for Unit-1 show no correlation indicating influx of
carbonates did not affect organic matter preservation. (b) Crossplot between Ca and ������ for Unit-2
showing a negative linear correlation, which indicatcates influx of carbonates especially detrital dolomites
resulted in dilution of organic matter.

The principal component analysis (PCA) described in Section 3.4.2 suggested that a major fraction
of Si in UBS is of biogenic or authigenic origin; therefore, I used Si concentration to study the effect of
dilution by biogenic sediments on OM richness. As shown in Figure 4.3, Unit-1 has a higher concentration
of Si in comparison to Unit-2. For Unit-1, Si shows no correlation with detrital proxying elements Al and
Ti (Figure 4.8), which indicates that a major fraction of Si in Unit-1 is associated with biogenic and
authigenic sources. Moreover, Figure 4.3 shows that Unit-1 consists of radiolarian-rich siliceous mudstone
(F1), which confirms the presence of biogenic and authigenic Si in this unit. However, Si shows a negative
correlation with ������ for Unit-1 (Figure 4.8c), which indicates that the biogenic silica associated with
radiolarians had a dilution effect in the accumulation of OM. The negative correlation between the detrital
clay proxying elements of Al and Ti with biogenic Si for Unit-1 samples indicates that biogenic silica also
had a dilution effect on the detrital clay influx. However, for Unit-2, Si correlates positively with Al and
Ti, which indicates that the Si of Unit-2 has detrital origin. This is also supported by the absence of siliceous
mudstone (F1) in Unit-2 (Figure 4.3). The Zr/Rb ratio, which is a proxy for the grain size of detrital
sediment, increases in Unit-2, suggesting an increase in the coarser or silt-sized fraction of the detrital
sediments in Unit-2. Therefore, for Unit 2, the absence of siliceous mudstone (F1) and an increase in the
Zr/Rb ratio indicate that much of the Si in Unit-2 is associated with silt-sized quartz grains. This is further
supported by the presence of silt content in macrofossil-bearing silt-rich mudstone (F4), which is
predominantly present in Unit-2 (Figure 4.3). Therefore, Si was used as proxy a for silt-sized siliciclastic
sediments for Unit-2. The crossplot of Si and ������ for Unit-2 samples shows a negative correlation,
which indicates that siliciclastic silt influx had a dilution effect during OM accumulation in Unit-2.

134
Figure 4.8: For Unit-1, crossplots between (a) Si and Al, and (b) Si and Ti, showing no covariance, which
indicates in Unit-1 has a considerable proportion of Si is associated with the radiolarians in siliceous
mudstone (F1). It also indicates that biogenic silica had a dilution effect on the detrital clay. (c) Crossplot
between Si and ������ for Unit-1 show a negative correlation indicating influx of biogenic Si had resulted
in the dilution of organic matter. For Unit-2, crossplot between (d) Si and Al (e) Si and Ti show positive
correlation indicating Si in Unit-2 has detrital origin. (f) Crossplot between Si and ������ for Unit-2
showing a negative correlation, which indicates influx of detrital Si associated with silt-sized sediment had
resulted in the dilution of organic matter.

Redox Proxy: C-S-Fe relationship

Figure 4.9a shows that the crossplot between total Fe (FeT) and total S (ST) shows a strong positive
correlation for all four UBS sub-units, indicating that Fe and S in the UBS are associated with pyrite.
However, Unit-1 samples plot approximately on the DOP=1 or the stochiometric pyrite line, while Unit-2
Fe vs. S plots below the stochiometric pyrite line. This indicates pyrite formation in Unit-1 was not limited
by the availability of Fe or S, while for Unit-2, pyrite formation was sulfur-limited. The ������ vs. S plot
in Figure 4.9b shows that all four sub-units have a positive covariance and the regression fit has a non-zero
intercept, indicating that UBS deposition took place in reducing conditions and not in the normal
oxygenated marine condition. However, the following trends are also visible from the plot in Figure 4.9b:
1) for Unit-1, ������ > .5 wt. % and total S varies between 2-4 wt.%; and 2) for Unit-2, ������ <
.5 wt% and S vary between 0.5-2 wt.%. The trend-line for Unit-1 has a higher slope than that of Unit-2,

135
which indicates that sulphide formation increased substantially after a threshold TOC value of around
12.5%. Therefore, for Unit-1, sulfate reduction increased rapidly due to the higher availability of OM for
degradation. This in turn increased the sulfide concentrations in the bottom waters, which resulted in a
persistent euxinic condition in Unit-1. However, the trend-line for Unit-2 samples suggests less OM
degradation has resulted in comparatively less sulfate reduction during the deposition of this unit, which
gave rise to a less reducing condition for Unit-2.

Figure 4.9: (a) Crossplot between total iron (FeT) and total sulfur (ST) for different chemostratigraphic sub-
units of UBS is shown. ST/FeT stochiometric ratio of 1.15 represents pyrite shown by the black solid line in
the plot, and the equivalent degree of pyritization (DOP) is equal to 1 for this line, which indicates the
euxinic condition. The green broken line represents regression fit for Unit-1and black broken line represents
the regression fit for Unit-2. Unit-1 plots close to DOP=1 line indicating that most of the iron in Unit-1 is
associated with pyrite and this unit was deposited in euxinic condition. Whereas Unit-2 plots below the
DOP=1, which indicates that supply of iron was more compared to sulfur and pyrite formation was sulfur
limited. This indicated Unit-2 was deposited at comparatively less reducing conditions than Unit-1. (b)
Crossplot between ������ and total sulfur (ST) for the different chemostratigraphic units of UBS is shown.
Solid line represents S and C trend for oxygeneted normal marine conditions during organic matter
deposited. Broken lines represent linear regression data for Unit-1 and Unit-2. The non-zero intercept of
the linear regression lines for both Unit-1 and Unit-2 in the crossplot indicates that their deposition has
taken place in reducing condition and not in oxygeneted condition.

Figure 4.10 shows the overlay of the UBS samples on a Fe-S-������ ternary diagram. This ternary
diagram was used for DOP evaluation, which provides further clarity on the paleoredox condition of various
sub-units of the UBS. Figure 4.10 shows that Unit-1 plots along a line that intersects the Fe-S axis at about
0.52, which is close to 0.54, where the DOP=1 line intersects the Fe-S axis. This suggests Unit-1 was
deposited in an euxinic paleoredox condition in a strongly stratified water column, as described in Section
4.3.4. Moreover, an intersect of 0.52 at the Fe-S axis is equivalent to a DOP of 0.96, which suggests that

136
96% of the Fe in Unit-1 is associated with pyrite. Samples for sub-unit 2a plots along a line that intersects
the Fe-S axis at approximately 0.30, which is equivalent to a DOP of 0.56; thus 56% of the iron in this sub-
unit is associated with pyrite. However, sub-unit 2b plots between DOPs of 0.56 and 0.75, which indicates
that the paleoredox condition fluctuated from suboxic to anoxic with an increasing intensity of anoxia
compared to that observed for sub-unit 2a.

Figure 4.10: Fe-S-TOCo g ternary diagram for UBS samples from well RT showing sub-unit 1a and 1b
mostly plot along DOP~0.96, which indicates a strongly euxinic condition for these sub-units. Sub-unit 2a
plots along DOP~0.56 suggesting suboxic condition for this sub-unit, while sub-unit 2b plots between DOP
of 0.56-0.75, suggesting for 2b the paleoredox condition fluctuated from suboxic to anoxic with increasing
intensity of anoxia compared to that in sub-unit 2a.

TOC-Trace Elements Relationship

Tribovillard et al. (2006) demonstrated that the covariance patterns between the concentration of
redox-sensitive trace elements (TE) with TOCo g can be used as a proxy to distinguish between various
paleo-redox conditions. The redox-sensitive TE, Mo and U, used in this study show an overall positive
correlation with TOCo g in the scatterplots, as shown in Figure 4.11. However, this figure also shows the
following three group of samples associated with the covariance pattern between TOCo g and Mo, and
TOCo g and U: 1) Group 1 is characterized by TOCo g <5 wt.%, no enrichment of Mo and limited
enrichment of U; 2) Group 2 has TOCo g s between 5-12.5 wt.%, and Mo and U are moderately enriched
and show a good linear correlations with TOCo g; 3) Group 3 has TOCo g>12.5 wt.%, and shows highly

137
enriched Mo and U concentrations, which increase exponentially with increase in TOCo g . Figure 4.11c
(modified from Tribovillard et al., 2006) shows the expected pattern of relative enrichment of Mo and U at
suboxic, anoxic and euxinic conditions. Therefore, comparing Figure 4.11a, b with Figure 4.11c helped in
distinguishing between the different redox conditions for UBS sub-units.

Index
Unit 2b
Unit 2a
Unit 1b
Unit 1a

Figure 4.11: Crossplot between (a) ������ vs Mo; and (b) ������ vs U showing three distinct covariation
patterns. (c) Chart showing the expected pattern of relative enrichment of Mo and U at suboxic, anoxic and
euxinic conditions (modified from Tribovillard et al., 2006), which suggest in suboxic condition, ������ <
5 wt.%; Mo has no enrichment while U shows limited enrichment. In anoxic condition, 5 < ������ <
.5 wt.%, both Mo and U shows moderate enrichment and has a good linear covariance pattern with
������ . In euxinic condition, ������ > .5 wt.%, concentration of Mo and U increases exponentially
and shows poor correlation with ������ .

As discussed in Section in 4.3.5, Mo enrichment starts only in the presence of dissolved HS − or a

suboxic paleoredox condition. However, in the suboxic paleoredox condition, limited enrichment of U can
take place as it does not require the dissolved of HS − to start the enrichment process. Therefore, during the

138
deposition of Group 1 samples, the paleoredox condition was suboxic, which explains no enrichment of
Mo but the limited enrichment of U for these samples. Figure 4.11 shows that Group 1 consists of samples
from the upper interval of sub-unit 2a, which suggests for 2a close to the 2b-2a boundary, the paleoredox
condition was suboxic. For this interval of sub-unit 2a, Mo ranges from 2-6 ppm, which is close to the
background levels of 1-5ppm found in average shale (McLennan, 2001). These observations suggest that
the upper sub-unit 2a was deposited in a suboxic benthic condition and dissolved sulfide was absent even
in the pore water below the sediment-water interface.

For Group 2 samples, the moderate increase in Mo and U concentrations suggests that the
paleoredox conditions were comparatively more reducing than those of Group 1 samples due to an increase
in the dissolved HS − concentration. Therefore, Group 2 samples were deposited in anoxic paleoredox
conditions. Mo and U show good linear correlation with ������ , as their uptake is dependent on the
availability of suitable organic matter (Tribovillard et al., 2006). Based on this analysis, it was interpreted
that the threshold 5 wt.% ������ is the boundary between the suboxic (Group 1) and anoxic (Group 2)
conditions for deposition of the UBS. Figure 4.11 shows that Group 2 is dominated by samples from sub-
unit 2b and the lower interval of sub-unit 2a, which suggesting these two group of samples were also
deposited in anoxic conditions. The bulk concentration of Mo for sub-unit 2b varies between 9-200 ppm,
and according to Scott and Lyons (2012), this suggests predominantly anoxic conditions with either an
intermittent euxinia or an increase in the degree of restriction in the basin. Lower sub-unit 2a samples (near
the 1b-2a boundary) were also deposited in anoxic conditions; however, these sub-unit 2a samples have
bulk Mo concentrations between 5-24 ppm, which are higher than the Mo concentration in average shale
(~1.3 ppm). Scott and Lyons (2012) suggested that an Mo enrichment between the crustal average of 1-
2ppm (Wedepohl, 1991) and 25ppm indicates the presence of dissolved HS − that is restricted within the
pore water; this suggests the euxinic conditions existed only below the sediment-water interface. This
suggests that deposition of sub-unit 2a started in an environment where the benthic redox conditions along
the sediment-water interface may have remained suboxic to anoxic, while euxinic conditions existed only
within the pore water below the sediment-water interface.

For Group 3, both Mo and U show exponential increase in concentrations and their correlation with
������ disappears. Exponential enrichment of Mo is explained by the authigenic uptake and 2-3 times
increased burial rate of Mo along with the sulfide minerals, which is associated with very high dissolved
HS − concentration (Scott and Lyons, 2012). High dissolved HS − concentration indicates a euxinic
paleoredox condition during the deposition for Group 3 samples. Mo resides primarily in pyrite, which
results in a weak covariance of Mo with ������ (Tribovillard et al., 2004). This explains the poor
covariance between Mo and ������ for the Group 3 samples in Figure 4.11. In euxinic conditions, U

139
resides in authigenic uptake consisting of metal sulfides and oxyhydroxides rather in the OM hosts, which
explains the poor correlation between U with ������ . Therefore, the threshold ������ of 12.5 wt.% marks
the limit between anoxic and euxinic conditions for the deposition of UBS near Well RT in Williston Basin.
Figure 4.11 shows that Group 3 is dominated by sub-units 1a and 1b, which indicates that these two sub-
units were deposited in an euxinic condition. The bulk concentration of Mo for these sub-units varies
between 100-400 ppm. Bulk concentration of Mo more than 100 ppm suggests the presence of dissolved
hydrogen sulfide in the water column throughout the year (Scott and Lyons, (2012). This further
substantiates the postulation of persistent euxinic conditions above the sediment-water interface for sub-
units 1a and 1b.

EF-Mo vs. EF-U relationship

Enrichment factors (EF) of U and Mo for the chemostratigraphic sub-units of UBS exhibit three
different clusters in the U-EF vs. Mo-EF plot for Well RT (Figure 4.12). For the first cluster, which is
dominated by the samples from Sub-units 1a and 1b, Mo-EF is 350-900 and U-EF is 70-150 (Figure 4.12).
Most Sub-unit 1a and 1b samples plot in the zone representing sulfurized OM, which is an extension of the
euxinic field characterized by a Mo/U ratio that is 1-3 times of that in modern seawater. The second cluster,
which is dominated by samples from both 2a and 2b, exhibit U-EFs of 20-30 and Mo-EFs of 20-100. As
shown in Figure 4.12, this cluster of 2a and 2b samples plot in the zone between the anoxic and euxinic
fields characterized by a Mo/U ratio that is 0.3-1 times of that of modern seawater. The third and last cluster,
which consists of samples from 2a, exhibit U-EFs of 10-20 and Mo-EFs of 3-10. As shown in Figure 4.12,
this cluster of 2a samples plot in the zone between suboxic and anoxic fields characterized by a Mo/U ratio
that is 0.1-0.3 times of that of modern seawater. The crossplot between EF-U and EF-Mo (Figure 4.12)
shows that organic matter in Sub-units 1a and 1b have been sulfurized. Similar, trend is shown by sulfurized
Late Jurassic Kashpir oil shales of the Russian Platform (Tribovillard et al., 2012). Sulfurization of organic
matter is described as the incorporation of reduced inorganic sulfur species or polysulfides into organic
matter by reaction with functionalized organic molecules (Tribovillard et al., 2004). Sulfurization of organic
matter takes place when hydrogen sulfide is formed in quantities exceeding the portion that can be
scavenged rapidly by highly reactive iron to form pyrite. Tribovillard et al. (2004) have demonstrated that
presence of sulfurized organic matter enhances Mo authigenic enrichment such that Mo is systematically
more enriched relative to the other redox-sensitive/sulfide-forming elements studied (U, V, Ni and Zn).
Thus, sulfurization indicates a strongly euxinic condition in the water column. In such euxinic conditions,
Mo is rapidly sequestered by sulfurized organic matter and also goes into solid solution with sulfide
minerals, so availability of both the host-phases in very high amount results in very high enrichment of Mo
relative to U as observed in the unit 1a and 1b samples of UBS. Raven et al. (2016) demonstrated that

140
organic matter is rapidly sulfurized while sinking through the water column in the semi-restricted Cariaco
Basin. They observed water column sulfurization is most extensive during periods of high primary
productivity and appears to involve elemental S, possibly via polysulfides. They also suggested that the
process has the potential to deliver substantial amounts of OM to the sediment by making it less available
for remineralization in the water column; this, in turn, increases the preservation potential of organic matter.

Figure 4.12: Crossplot between U Enrichment factors (U-EF) vs. Mo Enrichment factors (Mo-EF) showing
three clusters of the UBS samples from core RT. The broken lines represent 0.3, 1 and 3 times of the average
Mo/U of present-day seawater (SW). The general patterns of U-EF vs. Mo-EF variation in modern marine
environments are represented by (i) gray field for unrestricted marine trend for eastern tropical Pacific
Ocean, (ii) green field for semi-restricted Cariaco Basin, where particulate shuttle function for Mo
enrichment within the water column and (iii) black solid line for strongly restricted Black Sea (adopted
from Tribovillard et al., 2012). The yellow field represents areas, where the presence of sulfurized organic
matter leads to very high enrichment of Mo in strongly euxinic conditions (Tribovillard et al., 2012). Sub-
unit 1a and 1b plots in the zone of sulfurized organic matter (yellow field), whereas sub-unit 2b and part of
sub-unit 2a and plots in the anoxic to the euxinic field (gray area). Finally, rest of the samples from sub-
unit 2a plot in the sub-oxic to the anoxic field in the gray area.

Mo-TOC Relationship and Basin Restriction

In anoxic to euxinic settings, the slope (m) of the regression line in Mo vs. TOC crossplot can be
used as a proxy to determine the degree of restriction in the basin. As shown in Figure 4.13, the value of m
for sub-units 1a, 1b and 2b are 21, 23.5 and 19.3, respectively. These m values are close to those of the
modern day Cariaco Basin (Mo/TOC value ~25), which is a moderately restricted basin (Algeo and Lyons,
2006; Algeo et al., 2007; Algeo and Rowe, 2012). This suggests that sub-units 1a, 1b, and 2b were deposited
in a semi-restricted basin, and, therefore, the water was subjected to intermittent deepwater renewal.
However, since the value of m is higher in sub-unit 1a than in sub-unit 1b, the basin was progressively
becoming less restricted due to sea level rise. The value of m decreases in Unit-2b, which indicates that the

141
basin became slightly more restricted during the deposition of sub-unit 1b. However, m value cannot be
used as a proxy for sediments deposited in oxic-suboxic conditions. Therefore, m for sub-unit 2a, which
was deposited in suboxic conditions, has no physical significance. Based on the procedure used by Algeo
et al., (2007) for calculating the deepwater water renewal time in Devono-Carboniferous silled basins using
m, the deepwater renewal time for the Williston Basin was estimated to be around 25 years (Figure 4.13b).

Figure 4.13: (a) Mo-TOCorg crossplot for UBS in Well RT showing slope (m) for the regression line for
chemostratigraphic sub-unit 1a, 1b and 2b are 21.1, 23.5 and 19.3 respectively. Value of m for these UBS
sub-units is close to that of the semi-restricted Cariaco Basin, which m value close to 25. (b) m versus
deepwater renewal times for modern anoxic silled basins showing an average m of 21 for UBS suggests the
deepwater renewal in the Williston Basin during the deposition of UBS was around 25 years (modified
from Algeo and Lyons, 2006).

Redox-condition and OM preservation

Different geochemical proxy based analyses show consistent results for the paleoredox condition
of the chemostratigraphic units of the UBS. These paleoredox conditions identified from the inorganic
geochemical analysis is consistent with the petrographic observations. Table 4.3 lists the findings for the
paleoredox conditions based on inorganic geochemical methods and petrographic observations. These
results indeed suggest that the redox-condition was changing during the deposition of the UBS. The bottom
water conditions were strongly euxinic, which intermittently extended to the photic zone during the
deposition of sub-unit 1a and 1b and this helped in the enhanced preservation of organic matter. However,
during the deposition of sub-unit 2a, the benthic condition became less reducing and varied from suboxic
to anoxic conditions. Suboxic conditions along the sediment-water interface resulted in more degradation
of organic matter, which in turn decreased the total amount of preserved organic matter. This is one of the
reasons for the low TOC in sub-unit 2a shown in Figure 4.3. However, as conditions became more anoxic

142
during the deposition of sub-unit 2b, the preservation potential of organic matter increased which resulted
in the high TOC in Unit-2b. Therefore, the variation in TOC in the UBS is strongly dependent on bottom
water redox-conditions.

Proxy for Paleoproductivity

As described in Section 4.3.9, I used the stable isotope nitrogen (δ15Norg) as a proxy for the
paleoproductivity in the UBS. Results of δ15Norg for the different chemostratigraphic units of the UBS of
Well RT are shown in Figure 4.14a. Figure 4.3 shows the δ15Norg profile, which suggests that δ15Norg has
varied between Unit-1 and Unit-2 of the UBS in Well RT. The TOCorg vs. δ 5No g crossplot in Figure
4.14b shows sub-units 1a and 1b have depleted average δ15Norg of around 1.9; Sub-unit 2a has a much
heavier δ15Norg of 6.9. The depleted values of δ15Norg in sub-units 1a and 1b suggest that the nutrient supply
in the surface water was high; this suggests increased paleoproductivity in the surface water during the
deposition of these sub-units. However, the relatively heavier values of δ15Norg in sub-unit 2a suggests an
increase in the utilization of the N15 isotope bearing nitrate, which was probably caused by a reduction in
nutrient supply. This reduction in surface water nutrient supply resulted in a decrease in primary
productivity during the deposition of sub-unit 2a. The δ15Norg values in Sub-unit 2b are slightly depleted
compared to those in sub-unit 2a, suggesting an increase in nutrient supply and thereby an increase in
productivity during the deposition of sub-unit 2b.

Figure 4.14: (a) Box and whisker plot showing the results for the stable isotope of nitrogen (δ 5No g for
the chemostratigraphic units of UBS in well RT. The mean values of δ15Norg for each chemostratigraphic
units are also labeled. (b) TOCorg vs δ 5No g crossplot showing that sub-unit 1a and 1b have high TOC and
lighter values of δ 5No g, whereas sub-units 2a has comparatively lower TOC and heavier values of
δ 5No g. Sub-unit 2b has mixed values of δ 5No g. This crossplot indicates paleoproductivity during the
deposition of Unit-1 was comparatively higher than that during the deposition of sub-unit 2a. The
paleoproductivity increased marginally again during the deposition of sub-unit 2b.

143
Table 4.3: Findings for the paleoredox conditions based on inorganic geochemical methods and petrographic observations.

Absolute Mo
Sub-units C-S-Fe relationship TOC-TE Relation EF-Mo vs EF-U Petrographic Evidence
Concentration

Primarily suspension
Exponential increase in Mo and Extreme enrichment of Mo
DOPT >0.96 sedimentation
U with TOCorg, weak relative to U Mo >100ppm
Euxinic Presence of planktonic
1a, 1b covariance Sulfurization of OM Persistent euxinia in
Strongly stratified fecal pellets (Section
TOCorg >12.5wt.% Strongly euxinic, photic bottom water
Water Column 3.4.1)
Euxinic zone euxinia
Absence of bioturbation

Lower: Moderate enrichment,

Upper: No to limited
DOPT <0.52 enrichment
Suboxic Lower: good linear covariance;
2a
Weakly stratified Upper: no covariance
water column Lower 5.0<TOCorg <12.5wt.%,
Upper<5wt.%
Lower: Anoxic, Upper: suboxic
0.75 <DOPT < 0.52 Moderate enrichment
Suboxic to anoxic Good linear covariance
2b
Weakly stratified 5.0<TOCorg <12.5wt.%,
water column Anoxic
Lower: Moderate Lower: 5<Mo< 25ppm,
enrichment of both U and Upper: Mo< 5ppm
Presence of agglutinated
Mo Upper: More Lower: Euxinic restricted
benthic forams (Section
enrichment of U than Mo to pore water, Upper:
3.4.1)
Lower: Anoxic to Euxinic, suboxic even in sediment
Upper: suboxic to anoxic column
Mo: 9-100ppm
Moderate enrichment of Anoxic to intermittent Occasional presence of
both U and Mo euxinia agglutinated benthic
Anoxic to Euxinic Basin getting restricted forams (Section 3.4.1)
(m~19)

144
4.4 Discussion

In this section, a conceptual model for accumulation and preservation of OM in the UBS is
proposed. This model provides the basis for the organic richness of four chemostratigraphic units of the
UBS with respect to the interplay between OM production, preservation, and dilution. The framework of
this conceptual model includes the effect of various depositional settings of the basin, such as basin
morphology, paleogeography, paleoclimate, and probable nutrient sources. Therefore, to build a framework
for the conceptual model, this section begins with selecting fit-for-purpose options for these depositional
settings in the context of the UBS.

Depositional Setting

Basin Morphology
During the deposition of the UBS in Early Mississippian, Williston Basin was a very gently sloping
epicontinental sea, which extended over hundreds of miles (Smith and Bustin, 1998). During the deposition
of the UBS, the basin was located near the equator and had a maximum depth of around 150m (Section
3.5.3). As discussed in Section 4.3.7, average m in Mo vs. TOCorg crossplot is 21, which indicates that the
basin was semi-restricted during the deposition of the UBS. This finding is aligned with the physiographic
presence of the Central Montana Uplift (Smith and Bustin, 1998), which formed a marginal sill to the west-
southwest and limited frequent deepwater renewal (Figure 4.15). Moreover, as suggested by Smith and
Bustin (1998), the Montana Trough along the south-western edge formed a connection between the
Williston Basin with the open ocean conditions at the western margin of the North American Craton. It is
proposed that this physiographic feature aided in the intermittent renewal of deep waters.

Paleoclimate
During the deposition of the UBS, the Williston Basin was characterized by a warm, tropical
climate with seasonal precipitation (Ettensohn and Barron, 1981). In the Early Mississippian, Williston
Basin was located within the tropics in the southern hemisphere and was under the influence of prevailing
east-west trade winds, as shown in Figure 4.15 (Algeo et al., 2007; Blakey, 2005). However, due to the
absence of any orographic barriers, the trade winds passed freely to the west and contributed very little
towards precipitation (Smith and Bustin, 1998). As the Intertropical Convergence Zone (ITCZ) was also
close to Williston Basin, the annual migration of the ITCZ probably caused seasonal rainfall, and a seasonal
change in the direction and intensity of the trade wind (De Vleeschouwer et al., 2014).

145
Figure 4.15: Schematic diagram showing the prevailing depositional setting of the Williston Basin during
the deposition of UBS. The basin was under the influence of the east-west trade wind as shown by the
yellow arrow due to its location near the equator. The Williston Basin was semi-restricted in nature due to
the presence of the Central Montana Uplift, which was situated at the south-west of the basin. Stratified
water column developed due to the formation of a thermocline, which stabilized the chemocline, which in
turn developed as a result of degradation of organic matter. The chemocline created a stagnant euxinic
bottom water condition during the deposition of Unit-1 of UBS. However, when the depth of the
thermocline is lowered due to fall in the sea level, the seasonal fluctuation of the thermocline disturbed the
chemocline. This, in turn, decreased the reducing condition at the sediment-water interface during the
deposition of Unit-2. The primary source of the nutrients was through regeneration of the biolimiting
nutrients N and P in euxinic condition.

Stratification of Water Column

Ettensohn and Barron (1981), and Lineback and Davidson (1982) suggested that a stratified water
column existed during the UBS deposition, which resulted in restricted circulation and mixing of surface
water with the deeper bottom waters. I confirmed the presence of a stratified water column based on the
DOPT values for the UBS which are more than 0.75 (see Section 4.3.4). Stratification may form in a water
column due to a difference in the density of surface and deeper water, which in turn may be caused by high
fluvial discharge and surface run-off into the basin. Smith and Bustin (1998) suggested that fluvial run-off
was minimum during the deposition of the UBS; therefore, the development of the stratified water column
due to density differences is an unlikely scenario. Stratification in the water column may also be caused by
temperature differences between the surface water layer with the deeper water. Tyson and Pearson (1991)
suggested that a thermocline may develop in modern shelf seas and ancient epeiric seas in temperate
climatic conditions due to seasonal variations in wind strength and temperature differences. However, the

146
development of thermocline is also prevalent in the large lakes located in tropical climatic conditions
(Lewis, 1996). Development of a thermocline stratifies the water column into three distinct layers: the
surface mixed layer, the intermediate layers along the thermocline, and the deeper bottom water (Tyson
and Pearson, 1991). These researchers suggested that in deep water areas in an epeiric sea, the water mass
stratification can be so stable that it results in the formation of meromictic conditions. In such conditions,
the deep bottom water resists seasonal overturn due to its higher volume and thus forms an isolated, deep,
stagnant pool. In meromictic conditions, density differences in the stratified water column due to
development of a thermocline eventually facilities chemical differences between the surface water and the
deep stagnant pool, which may result in the formation of a redox-boundary or chemocline in the deeper
bottom water (Boehrer and Schultze, 2008). However, seasonal fluctuation of the thermocline and surface-
water mixing results in fluctuation in the depth of the chemocline in the water column (Parkin and Brock,
1981). Therefore, it is proposed that the proximity of Williston Basin to the paleo-equator in a warm tropical
climatic setting, and a seasonal variation in direction and intensity of the trade wind resulted in the
development of a thermocline during the deposition of the UBS, which prevented vertical mixing of the
surface water with the deep bottom water (Figure 4.15). Moreover, the physiographic sill of the Central
Montana Uplift limited frequent deepwater exchange and thus did not disturb the deeper water. Therefore,
for the UBS, the deposition of OM matter resulted in rapid depletion of oxygen content in the deep stagnant
pool, which facilitated the development of a chemocline within the deeper water (Figure 4.15).

Paleoproductivity
Smith and Bustin (1998) suggested that during the deposition of the UBS, nutrients were supplied
through upwelling and in-situ regeneration (of nutrients) by the process of remineralization of OM.
Upwelling of nutrient-rich equatorial undercurrents was happening in the Central Montana Trough in an
estuarine-like circulation pattern occurring above a stagnant anoxic pool. An estuarine-like circulation
pattern generally takes place in environments with humid climatic conditions, which contribute to high
river discharge and thus result in a stratified water column due to salinity differences (Algeo et al., 2008).
However, during the deposition of the UBS, Williston Basin was characterized by a warm tropical climate
with only seasonal precipitation and low fluvial discharge; therefore, an estuarine-like circulation pattern
probably did not exist. Moreover, as the maximum depth of the basin was about 150 m, it is unlikely that
the equatorial currents in the Pacific Ocean, which according to Jewell (1995) exist at depths of more than
100-200m, would have entered the shallow epicontinental sea of Williston Basin. Therefore, it is postulated
that regeneration of biolimiting nutrients, such as P and N, was a major source of nutrients during the UBS
deposition. In euxinic conditions, P and N are preferentially released from OM, which then diffuses upward
from the sediment into a water column, and thus aids in the efficient recycling of nutrients (Ingall and
Jahnke, 1994, 1997). Periodic mixing of the surface and mid-water along the thermocline may result in

147
fluctuation of the oxic-anoxic and anoxic-euxinic chemocline in the deeper stagnant water; this, in turn,
probably aids in releasing the regenerated nutrients back to the surface water (März et al., 2008). In oxic-
suboxic conditions, P is entrapped in sediment as a result of redox cycling of Fe with P (Tribovillard et al.,
2006; März et al., 2008). This results in a decreased nutrient supply in the surface water, which in turn
decreases surface water productivity. Therefore, based on the redox condition of the various
chemostratigraphic sub-units of the UBS, P and N will either get preferentially released or will be entrapped
within the sediments. As shown in Figure 4.15, for UBS chemostratigraphic sub-units, the preferential
release will enhance the nutrient supply in the surface water during euxinic conditions, thereby increasing
paleoproductivity. Smith and Bustin (1998) suggested that most of the detrital sediment in the Bakken are
windblown, supplied from the Canadian Shield by the east-west trade winds. Therefore, along with in situ
regeneration of nutrients, it was hypothesized that during the deposition of UBS, the influx of clay into the
Williston Basin from the Canadian Shield by the northeast trade winds brought Fe-rich nutrients, which
augmented the OM productivity (Figure 4.15). This hypothesis is based on the positive covariance between
TOCorg and detrital proxying elements (Al and K) and was discussed in Section 4.3.3. Filipsson et al. (2011)
proposed a similar postulation for increased OM productivity on the northwest coast of Africa, due to the
influx of Saharan eolian dust, which bring Fe-rich nutrients in the Atlantic Ocean.

Model for Organic Richness in Chemostratigraphic Units of UBS

Sub-unit 1a
Deposition of the UBS started with the onset of transgression/flooding of Williston Basin from the
west (Section 4.4.1). This marks the beginning of the deposition of sub-unit 1a, which is the lower-most
chemostratigraphic sub-unit of the UBS. Availability of more nutrients due to transgression enhanced the
primary productivity, which in turn resulted in a higher flux of OM to the basin floor. This initiated the
development of anoxic benthic conditions due to degradation of OM. Accumulation of OM at the basin
floor continued incessantly, which exhausted the oxygen content in the bottom water and the reducing
condition continued to get intense. Consequently, the concentration of dissolved HS- within the bottom
water increased due to sulfate reduction, and as a result, the paleoredox condition at the basin floor became
euxinic. This resulted in the formation of a redox-boundary or chemocline in the bottom water, as shown
in Figure 4.16a. For sub-unit 1a, persistently euxinic conditions in the bottom water with intermittent photic
zone euxinia was confirmed by the geochemical redox-proxies and petrographic data (Table 4.3).
Moreover, as previously described, thermal stratification in the water column prohibited the mixing of the
sulfidic deep water with the oxygenated surface water, which helped in maintaining a stagnant euxinia at
the bottom water. In such euxinic conditions, the biolimiting nutrients (P, N) were preferentially released
from OM. As described in Section 4.4.1, a seasonal variation in the depth of the thermocline resulted in a

148
fluctuation in the depth of the chemocline, which aided in replenishing the surface water with the
regenerated nutrients. Therefore, the surface water remained oxygenated and well-supplied with nutrients,
which is evident from the depleted values of δ15Norg in sub-unit 1a, resulting in its high organic productivity.
This claim is further supported by the presence of planktonic fecal pellets in massive to finely laminated
facies (F2) and the abundance of radiolarians in siliceous mudstone (F1) in sub-unit 1a. Therefore, high
OM productivity along with high preservation potential due to bottom water euxinic conditions resulted in
increased organic richness in this sub-unit. Sulfurization of OM during intermittent photic zone euxinia
may have also helped in the enhanced preservation of OM during the deposition of sub-unit 1a. As discussed
in Section 4.3.3, no covariance between detrital clay and carbonate proxying elements with TOC indicates
that sedimentation of these elements was low, which did not aid in the dilution of OM. However, a negative
correlation between Si and ������ indicates that biogenic silica associated with radiolarians in siliceous
mudstone resulted in the dilution of OM.

Sub-unit 1b
A flooding event marks the boundary between sub-units 1a and 1b as discussed in Section 3.5.5.
Therefore, during the deposition of sub-unit 1b, there was a further rise in the sea level (Figure 4.16b). As
a result, the basin became less restricted; this is also suggested by a higher value of slope (m) for sub-unit
1b in the TOC-Mo crossplot compared to that in 1a (Section 4.3.7). Due to rise in sea level, the height of
the oxygenated surface water column increased, which in turn increased the travel time of OM through the
oxygenated water (Figure 4.16b). Therefore, the export productivity was slightly lower during the
deposition of sub-unit 1b because more OM was degrading before it was able to reach the euxinic bottom
of the basin. This decrease in the export productivity of OM is confirmed by a slight decrease in the TOC
of sub-unit 1b, in comparison to that of sub-unit 1a. However, during the deposition of sub-unit 1b, the
primary productivity in the surface water remained the same, which is confirmed by the insignificant change
in nitrogen isotope data (Section 4.3.9). All geochemical redox-proxies indicate that both sub-units 1a and
1b were also deposited in euxinic conditions (Table 4.3). However, as shown in Figure 4.3, for sub-unit 1b,
Mo, U, and TOC show an increasing trend with depth. The decrease in Mo concentration either indicates
low availability of host-phase (OM) or a decreased source-ion (aqueous Mo) availability (Section 4.2.1).
However, a decrease in source ion-availability would be suggestive of a more restricted basin, which is not
the case, as discussed earlier. Therefore, the decreasing trend in Mo is associated with a decrease in the
availability of its host-phase or OM during the deposition of sub-unit 1b. The effect of the influx of detrital
clay and dilution by biogenic silica in controlling OM richness in sub-unit 1b was the same as that for 1a
(Section 4.3.3).

149
Figure 4.16: Schematic diagrams showing the depositional condition of the different chemostratigraphic sub-units of UBS at the location of well
RT. (i) Stratified water column shown developed during the deposition of sub-unit 1a due to the formation of a thermocline. Accumulation of OM
continued incessantly, which exhausted the oxygen content in the bottom water and thereby resulted in the development of a stagnant euxinic
condition below the chemocline. The biolimiting nutrients P and N were preferentially released in a euxinic condition, which resulted in high surface
water productivity. Seasonal fluctuation of the depth of the thermocline also resulted in fluctuation of the chemocline which aided in releasing the
nutrients back to the surface water. (ii) The relative rise in sea level during the deposition of sub-unit 1b increased the height of the oxygenated
surface water column which decreased the export productivity since more OM were degraded while settling down through the water column.
Primary productivity in the surface water was high. Bottom water conditions were still euxinic which extended to the photic zone intermittently.
(iii) The sea level drop resulted in disappearance of radiolarians, and the depth of the thermocline dropped due to the fall in relative sea level during
the deposition of sub-unit 2a. So, the thermocline became closer to the chemocline. Seasonal fluctuation of the thermocline destabilized the
chemocline due to the partial mixing of the water column. This resulted in the formation of sub-oxic to anoxic condition along the sediment-water
interface. Paleoproductivity decreased as the biolimiting nutrients were less efficiently regenerated in sub-oxic condition. As more land area got
exposed due to relative sea level fall, it was covered by vegetation which became the source of more terrestrial OM during the deposition of sub-
unit 2a. (iv) During the deposition of sub-unit 2b, there was a minor increase in the depth of the thermocline due to a slight increase in relative sea
level. As a result, the chemocline stabilized, and conditions were more reducing which enhanced the regeneration of the biolimiting nutrients (P
and N). Therefore, the surface water productivity was higher compared to that of sub-unit 2b.

150
Sub-unit 2a
The absence of siliceous mudstone (F1) from sub-unit 2a is one of the major factors which makes
it different from sub-units 1a and 1b. The absence of facies F1 in sub-unit 2a is associated with the
disappearance of radiolarians, which thrive in a deeper water setting and are sensitive to falls in sea level
(De Wever et al., 2002). Therefore, the absence of radiolarians from sub-unit 2a indicates that the deposition
of this sub-unit is associated with a relative fall in sea level (Figure 4.16c). As discussed in Section 4.3.3,
an increase in Zr/Rb ratio and Ca in sub-unit 2a supports the claim that the influx of detrital siliciclastic and
dolomite silt sediment increased. The relative fall in sea level during the deposition of sub-unit 2a exposed
more basin margin area, which resulted in an increased influx of silt-sized sediment due to weathering. This
increase in the influx of the silt-sized sediments resulted in a dilution of OM, which was evident from the
negative correlation in TOC- Si and TOC-Ca crossplots (Figure 4.6 and Figure 4.7). In contrast, the influx
of detrital clay helped in the preservation of OM, as discussed in Section 4.3.3. Fe-rich clay also acted as a
source of nutrients and thus enhanced the productivity, as discussed in Section 4.4.1. As more land area got
exposed due to relative sea level fall, it was covered by vegetation, which became the source for the
terrestrial Type III OM present in sub-unit 2a (Section 0).

As shown in Figure 4.16c, the depth of the thermocline dropped with the continued fall in sea level
during the deposition of sub-unit 2a. As thermocline moved close to chemocline, the deeper bottom water
below the chemocline, which was stagnant during the deposition of sub-units 1a and 1b, did not remain
stagnant anymore due to the disturbance caused by the seasonal variation of the thermocline. This resulted
in the partial mixing of the bottom water, which developed suboxic to anoxic conditions at the sediment-
water interface. This was confirmed by the geochemical redox-proxies and petrographic evidence, which
indicated that sub-unit 2a was deposited in predominantly suboxic to anoxic conditions (Table 4.3). An
increase in oxygen content in bottom water resulted in a less effective regeneration of nutrients like P,
which got retained within the sediment by getting co-precipitated with Fe-oxyhydroxides (März et al.,
2008). This, in turn, decreased the nutrient supply and, as a result, the surface water productivity diminished,
which was confirmed by the lower δ15Norg in sub-unit 2a in Figure 4.14. Suboxic conditions during the
deposition of sub-unit 2b resulted in an increase in OM degradation and thus decreased the preservation
potential at the bottom of the basin. Therefore, the decrease in TOC in sub-unit 2a is a combined effect of
low preservation due to suboxic paleoredox conditions, low productivity, and more dilution by detrital silt-
sized sediments and all three of these factors are related to a relative fall in sea level.

Sub-unit 2b
Geochemical redox proxies for sub-unit 2b indicate that benthic conditions were anoxic with
intermittent euxinia (Table 4.3). Therefore, the paleoredox conditions during the deposition of sub-unit 2b

151
were more reducing compared to those of sub-unit 2a. It is proposed that during the deposition of sub-unit
2b, there was a minor rise in the sea level, which in turn raised the depth of the thermocline and surface
water. As a result, the chemocline stabilized and thus increased the stagnation in the deeper bottom water
below the chemocline, making the conditions more reducing than those of sub-unit 2a. Moreover, the
intermittent euxinia resulted in a better regeneration of the biolimiting nutrients, thereby increasing the
surface water productivity, which is evident from the higher δ15Norg compared to that of sub-unit 2a (Figure
4.14). Dilution of OM by detrital silt-sized sediments was similar to that of sub-unit 2a, which is evident
from the negative correlation in TOC- Si and TOC-Ca crossplots (Figure 4.6 and Figure 4.7). Therefore,
organic richness in sub-unit 2b is a result of both enhanced paleoproductivity and better preservation
potential due to development of intermittent euxinic conditions at the sediment-water interface.

4.5 Conclusion

• Two chemostratigraphic units, namely Unit-1 and Unit-2 were identified in Well RT. Unit-1 was
further subdivided into two sub-units: 1a and 1b. Unit-2 was also subdivided into two sub-units: 2a
and 2b. The subdivision of the UBS was primarily based on the trends of detrital and biogenic
proxying elements and the variation in the response of the redox sensitive trace elements.
• The organic richness of the UBS at the location of Well RT or the southern part of the basin
depocenter was controlled by paleoproductivity, bottom-water redox conditions, and dilution by
detrital silt-size sediments and radiolarians. However, for the four sub-units of the UBS, these factors
varied, primarily due to a relative change in sea level, which in turn affected the organic richness of
these sub-units. Therefore, a variation in organic richness is observed in the four chemostratigraphic
sub-units of the UBS.
• The slope (m) of the Mo-TOC crossplot indicated that the basin was semi-restricted in nature due
to the presence of the Central Montana Uplift, which was located at the south-east of the basin. The
higher value of m in sub-unit 1b in comparison to that of 1a suggests that the basin was getting less
restricted due to relative sea level rise. However, increase in the detrital sediment influx during the
deposition of both sub-units 2a and 2b and the comparatively lower value of m during the deposition
of sub-unit 2b indicates that Unit-2 was deposited during a relative sea level fall.
• The degree of pyritization for Unit-1 indicated that a stratified water column existed in Williston
Basin during the deposition of sub-units 1a and 1b. However, the stratification was disturbed during
the deposition of Unit-2 due to fall in relative sea level and quasiperiodic fluctuation of the
thermocline.

152
 • Unit-1 was deposited in strongly euxinic conditions which existed in the bottom water and also
extended to the photic zone intermittently. During the deposition of Unit-2, the conditions were
predominantly sub-oxic to anoxic along the sediment-water interface.
• Paleoproductivity was high during the deposition of Unit-1 as more biolimiting nutrients (P, N) were
getting preferentially regenerated due to the euxinic paleoredox conditions. However, as the
paleoredox conditions became less reducing during the deposition of Unit-2, paleoproductivity
declined due to the decreased regeneration of the nutrients.

4.6 References

Algeo, T. J., P. H. Heckel, J. B. Maynard, R. Blakey, and H. Rowe, 2008, Modern and ancient epicratonic
seas and the superestuarine circulation model of marine anoxia: Dynamics of Epeiric Seas:
Sedimentological, Paleontological and Geochemical Perspectives: Geological Association of Canada
Special Publications, v. 48, p. 7–38.
Algeo, T. J., and T. W. Lyons, 2006, Mo-total organic carbon covariation in modern anoxic marine
environments: Implications for analysis of paleoredox and paleohydrographic conditions:
Paleoceanography, v. 21, no. 1.
Algeo, T. J., T. W. Lyons, R. C. Blakey, and D. Jeffrey Over, 2007, Hydrographic conditions of the
Devono–Carboniferous North American Seaway inferred from sedimentary Mo–TOC relationships:
Palaeogeography, Palaeoclimatology, Palaeoecology, v. 256, no. 1, p. 204–230.
Algeo, T. J., and J. B. Maynard, 2004, Trace-element behavior and redox facies in core shales of Upper
Pennsylvanian Kansas-type cyclothems: Chemical Geology, v. 206, no. 3–4, p. 289–318.
Algeo, T. J., and H. Rowe, 2012, Paleoceanographic applications of trace-metal concentration data:
Chemical Geology, v. 324–325, p. 6–18.
Algeo, T. J., and N. Tribovillard, 2009, Environmental analysis of paleoceanographic systems based on
molybdenum-uranium covariation: Chemical Geology, v. 268, no. 3–4, p. 211–225.
Altabet, M. A., and R. Francois, 1994, Sedimentary nitrogen isotopic ratio as a recorder for surface ocean
nitrate utilization: Global biogeochemical cycles, v. 8, no. 1, p. 103–116.
Altabet, M. A., and R. Francois, 1994, The use of nitrogen isotopic ratio for reconstruction of past changes
in surface ocean nutrient utilization, in Carbon cycling in the glacial ocean: constraints on the ocean’s
role in global change: Springer, p. 281–306.
Arthur, M. A., and B. B. Sageman, 1994, Marine Black Shales: Depositional Mechanisms and
Environments of Ancient Deposits: Annual Review of Earth and Planetary Sciences, v. 22, no. 1, p.
499–551.
Boehrer, B., and M. Schultze, 2008, Stratification of lakes: Reviews of Geophysics, v. 46, no. 2, p. 1–27.
Bohacs, K. M., A. R. Carroll, P. J. Mankiewicz, K. J. Miskell-Gerhardt, J. O. N. R. Schwalbach, M. B.
Wegner, and J. A. T. Simo, 2005, Production, destruction, and dilution-the many paths to source-rock
development: in N.B. Harris, ed., The Deposition of Organic-Carbon-Rich Sediments: Models,
Mechanisms, and Consequences: SEPM, Special Publication 82, p. 61–101.
Calvert, S. E., 1987, Oceanographic controls on the accumulation of organic matter in marine sediments:
Geological Society, London, Special Publications, v. 26, no. 1, p. 137–151.

153
Caplan, M. L., and R. M. Bustin, 1998, Palaeoceanographic controls on geochemical characteristics of
organic-rich Exshaw mudrocks: Role of enhanced primary production: Organic Geochemistry, v. 30,
no. 2–3, p. 161–188.
Chen, R., and S. Sharma, 2016, Role of alternating redox conditions in the formation of organic-rich interval
in the Middle Devonian Marcellus Shale, Appalachian Basin, USA: Palaeogeography,
Palaeoclimatology, Palaeoecology, v. 446, p. 85–97.
Dean, W. E., and M. A. Arthur, 1989, Iron-sulfur-carbon relationships in organic-carbon-rich sequences I:
Cretaceous Western Interior Seaway: American Journal of Science, v. 289, no. 6, p. 708–743.
Demaison, G. J., and G. T. Moore, 1980, Anoxic environment and oil source bed genesis: Organic
Geochemistry, v. 2, no. 8, p. 9–31.
Egenhoff, S. O., and N. S. Fishman, 2013, Traces in the dark-sedimentary processes and facies gradients in
the upper shale member of the Upper Devonian-Lower Mississippian Bakken Formation, Williston
Basin, North Dakota, U.S.A.: Journal of Sedimentary Research, v. 83, no. 9, p. 803–824.
Ettensohn, F. R., and L. S. Barron, 1981, Depositional model for the Devonian-Mississippian black-shale
sequence of North America: a tectono-climatic approach: T.C. Roberts, ed., Geological Society of
America, Cincinnati ′81 field trip guidebooks, Volume II, Economic geology, structure, American
Geological Institute, Falls Church, VA, p. 344-361.
Filipsson, H. L., O. E. Romero, J. B. W. Stuut, and B. Donner, 2011, Relationships between primary
productivity and bottom-water oxygenation off northwest Africa during the last deglaciation: Journal
of Quaternary Science, v. 26, no. 4, p. 448–456.
Hackley, P. C., and B. J. Cardott, 2016, Application of organic petrography in North American shale
petroleum systems: A review: International Journal of Coal Geology, v. 163, p. 8–51.
Hatch, J. R., and J. S. Leventhal, 1992, Relationship between inferred redox potential of the depositional
environment and geochemistry of the Upper Pennsylvanian (Missourian) Stark Shale Member of the
Dennis Limestone, Wabaunsee County, Kansas, U.S.A.: Chemical Geology, v. 99, no. 1–3, p. 65–82.
Hayes, M. D., 1985, Conodonts of the Bakken Formation (Devonian and Missippian), Williston Basin,
North Dakota: The Mountain Geologist, v. 22, no. 2, p. 64–77.
Helz, G. R., C. V. Miller, J. M. Charnock, J. F. W. Mosselmans, R. A. D. Pattrick, C. D. Garner, and D. J.
Vaughan, 1996, Mechanism of molybdenum removal from the sea and its concentration in black
shales: EXAFS evidence: Geochimica et Cosmochimica Acta, v. 60, no. 19, p. 3631–3642.
Ibach, L. E. J., 1982, The relationship between sedimentation rate and total organic carbon content in
ancient marine sediments: AAPG Geologisys Bulletin, v. 66, no. 2, p. 170–188.
Ingall, E., and R. Jahnke, 1994, Evidence for enhanced phosphorus regeneration from marine sediments
overlain by oxygen depleted waters: Geochimica et Cosmochimica Acta, v. 58, no. 11, p. 2571–2575.
Ingall, E., and R. Jahnke, 1997, Influence of water-column anoxia on the elemental fractionation of carbon
and phosphorus during sediment diagenesis: Marine Geology, v. 139, no. 1–4, p. 219–229.
Jaminski, J., T. J. Algeo, J. B. Maynard, and J. C. Hower, 1998, Climatic origin of dm-scale compositional
cyclicity in the Cleveland Member of the Ohio Shale (Upper Devonian), Central Appalachian Basin,
USA: Shales and mudstones, v. 1, p. 217–242.
Jarvie, D. M., 2012, Shale Resource Systems for Oil and Gas: Part 1—Shale-gas Resource Systems: Shale
reservoirs—Giant resources for the 21st century: AAPG Memoir 97, p. 69–87.

154
Jiang, C., M. Li, K. G. Osadetz, L. R. Snowdon, M. Obermajer, and M. G. Fowler, 2001, Bakken/Madison
petroleum systems in the Canadian Williston Basin. Part 2: Molecular markers diagnostic of Bakken
and Lodgepole source rocks: Organic Geochemistry, v. 32, no. 9, p. 1037–1054.
Jin, H., 2014, Source rock potential of the Bakken shales in the Williston Basin, North Dakota and Montana:
PhD Thesis, Colorado School of Mines, 220 p.
Jones, B., and D. A. C. Manning, 1994, Comparison of geochemical indices used for the interpretation of
palaeoredox conditions in ancient mudstones: Chemical Geology, v. 111, no. 1–4, p. 111–129.
Klinkhammer, G. P., and M. R. Palmer, 1991, Uranium in the oceans: Where it goes and why: Geochimica
et Cosmochimica Acta, v. 55, no. 7, p. 1799–1806.
Kocman, K. B., 2014, Interpreting Depositional and Diagenetic Trends in the Bakken Formation Based on
Handheld X-Ray Fluorescence Analysis, Mclean, Dunn, and Mountrail Countries, North Dakota: MS
thesis, Colorado School of Mines, 156 p.
Lash, G. G., and D. R. Blood, 2014, Organic matter accumulation, redox, and diagenetic history of the
Marcellus Formation, southwestern Pennsylvania, Appalachian basin: Marine and Petroleum
Geology, v. 57, p. 244–263
Lewis, W. M. J., 1996, Tropical lakes: how latitude makes a difference: in K. H. Timotius and F. Goltenboth
eds., Tropical limnology, v. 1, Springer, p 29–44.
Lineback, J. A., and M. L. Davidson, 1982, The Williston Basin–sediment-starved during the Early
Mississippian, in J. E. Christopher and J. Kaldi, eds., Fourth International Williston Basin Symposium:
Saskatchewan Geological Society Special Publication 6, p. 125-130.
Longman, M., K. Kocman, and L. Wray, 2014, Petrography of the Bakken black “shales” in the eastern
Williston Basin of North Dakota: AAPG Datapages/Search and Discovery Article #90193, Rocky
Mountain Section AAPG Annual Meeting, Denver, Colorado, July 20–22.
Lyons, T. W., and R. A. Berner, 1992, Carbon-sulfur-iron systematics of the uppermost deep-water
sediments of the Black Sea: Chemical Geology, v. 99, no. 1, p. 1–27.
Marchitto, T. M., 2007, Paleoceanography, Physical and Chemical Proxies/Nutrient Proxies Nutrient:
Encyclopedia of Quaternary Science, p. 1732–1740.
März, C., S. W. Poulton, B. Beckmann, K. Küster, T. Wagner, and S. Kasten, 2008, Redox sensitivity of P
cycling during marine black shale formation: Dynamics of sulfidic and anoxic, non-sulfidic bottom
waters: Geochimica et Cosmochimica Acta, v. 72, no. 15, p. 3703–3717
Maynard, J. B., 1981, Carbon isotopes as indicators of dispersal patterns in Devonian-Mississippian shales
of the Appalachian Basin: Geology, v. 9, no. 6, p. 262–265.
McLennan, S. M., 2001, Relationships between the trace element composition of sedimentary rocks and
upper continental crust: Geochemistry, Geophysics, Geosystems, v. 2, no. 4 (paper# 2000GC000109).
Meissner, F. F., 1978, Petroleum geology of the Bakken Formation, Williston Basin, North Dakota and
Montana, in D. Rehrig, eds., The economic geology of the Williston Basin: Proceedings of the
Montana Geological Society, 24th Annual Conference, p. 207-227.
Montoya, J. P., 1994, Nitrogen isotope fractionation in the modern ocean: implications for the sedimentary
record, Carbon cycling in the glacial ocean: constraints on the ocean’s role in global change: Springer,
Berlin, Heidelberg, p. 259–279.
Morford, J. L., W. R. Martin, and C. M. Carney, 2009, Uranium diagenesis in sediments underlying bottom
waters with high oxygen content: Geochimica et Cosmochimica Acta, v. 73, no. 10, p. 2920–2937.

155
Morford, J. L., W. R. Martin, R. François, and C. M. Carney, 2009, A model for uranium, rhenium, and
molybdenum diagenesis in marine sediments based on results from coastal locations: Geochimica et
Cosmochimica Acta, v. 73, no. 10, p. 2938–2960.
Murphy, A. E., B. B. Sageman, D. J. Hollander, T. W. Lyons, and C. E. Brett, 2000, Black shale deposition
and faunal overturn in the Devonian Appalachian basin: Clastic starvation, seasonal water-column
mixing, and efficient biolimiting nutrient recycling: Paleoceanography, v. 15, no. 3, p. 280–291.
Parkin, T. B., and T. D. Brock, 1981, The role of phototrophic bacteria in the sulfur cycle of a meromictic
lake: Limnology and Oceanography, v. 26, no. 5, p. 880–890.
Pedersen, T. F., and S. E. Calvert, 1990, Anoxia vs. productivity: what controls the formation of organic-
carbon-rich sediments and sedimentary rocks: AAPG Bulletin, v. 74, p. 454–466
Price, L. C., T. Ging, T. Daws, A. Love, M. Pawlewicz, and D. Anders, 1984, Organic Metamorphism in
the Mississippian-Devonian Bakken Shale, North Dakota Portion of the Williston Basin: in J.
Woodward, F. F. Meissner and J. L. Clayton, eds., Hydrocarbon Source Rocks of the Greater Rocky
Mountain Region, Rocky Mountain Association of Geologists, p. 83–133.
Raiswell, R., and R. A. Berner, 1985, Pyrite Formation in Euxinic and Semi-euxinic Sediments: American
Journal of Science, v. 285, p. 710–724.
Raiswell, R., F. Buckley, R. A. Berner, and T. F. Anderson, 1988, Degree of pyritization of iron as a
paleoenvironmental indicator of bottom-water oxygenation: Journal of Sedimentary Research, v. 58,
p. 812–819.
Raven, M. R., A. L. Sessions, J. F. Adkins, and R. C. Thunell, 2016, Rapid organic matter sulfurization in
sinking particles from the Cariaco Basin water column: Geochimica et Cosmochimica Acta, v. 190, p.
175–190.
Requeffo, A. G., J. Allan, S. Creaney, N. R. Gray, and K. S. Cole, 1991, Aryl isoprenoids and diaromatic
carotenoids in Paleozoic source rocks and oils from the Western Canada and Williston Basins: Organic
Geochemistry, v. 19, p. 245–264.
Rimmer, S. M., J. A. Thompson, S. A. Goodnight, and T. L. Robl, 2004, Multiple controls on the
preservation of organic matter in Devonian-Mississippian marine black shales: Geochemical and
petrographic evidence: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 215, no. 1–2, p. 125–
154.
Sageman, B. B., A. E. Murphy, J. P. Werne, C. A. Ver Straeten, D. J. Hollander, and T. W. Lyons, 2003,
A tale of shales: The relative roles of production, decomposition, and dilution in the accumulation of
organic-rich strata, Middle-Upper Devonian, Appalachian basin: Chemical Geology, v. 195, no. 1–4,
p. 229–273.
Scott, C., and T. W. Lyons, 2012, Contrasting molybdenum cycling and isotopic properties in euxinic
versus non-euxinic sediments and sedimentary rocks: Refining the paleoproxies: Chemical Geology,
v. 324–325, p. 19–27.
Scott, C., J. F. Slack, and K. D. Kelley, 2017, The hyper-enrichment of V and Zn in black shales of the Late
Devonian-Early Mississippian Bakken Formation (USA): Chemical Geology, v. 452, p. 24–33.
Smith, M. G., and R. M. Bustin, 2000, Late Devonian and Early Mississippian Bakken and Exshaw Black
Shale Source Rocks, Western Canada Sedimentary Basin: A Sequence Stratigraphic Interpretation:
AAPG Bulletin, v. 84, no. 7, p. 940–960.
Smith, M. G., and R. M. Bustin, 1996, Lithofacies and paleoenvironments of the Late Devonian and Early
Mississippian Bakken Formation, Williston Basin: Bulletin of Canadian Petroleum Geology, v. 44,
no. 3, p. 495–507.

156
Smith, M. G., and R. M. Bustin, 1998, Production and preservation of organic matter during deposition of
the Bakken Formation (Late Devonian and Early Mississippian), Williston Basin: Palaeogeography,
Palaeoclimatology, Palaeoecology, v. 142, no. 3–4, p. 185–200.
Sonnenberg, S. A., and A. Pramudito, 2009, Petroleum geology of the giant Elm Coulee field, Williston
Basin: AAPG Bulletin, v. 93, no. 9, p. 1127–1153.
Sonnenberg, S. A., C. Theloy, and H. Jin, 2017, The giant continuous oil accumulation in the Bakken
Petroleum System, Williston Basin: AAPG Memoir, v. 113, no. 110220, p. 91–120,
doi:10.1306/13572002M113508.
Stasiuk, L. D., 1993, Algal bloom episodes and the formation of bituminite and micrinite in hydrocarbon
source rocks: evidence from the Devonian and Mississippian, northern Williston Basin, Canada:
International Journal of Coal Geology, v. 24, no. 1–4, p. 195–210.
Tissot, B. P., and D. H. Welte, 1984, Diagenesis, catagenesis and metagenesis of organic matter, in
Petroleum Formation and Occurrence: Springer, p. 69–73.
Tribovillard, N., T. J. Algeo, F. Baudin, and A. Riboulleau, 2012, Analysis of marine environmental
conditions based on molybdenum-uranium covariation-Applications to Mesozoic paleoceanography:
Chemical Geology, v. 324–325, p. 46–58.
Tribovillard, N., T. J. Algeo, T. Lyons, and A. Riboulleau, 2006, Trace metals as paleoredox and
paleoproductivity proxies: An update: Chemical Geology, v. 232, no. 1–2, p. 12–32.
Tribovillard, N., A. Riboulleau, T. Lyons, and F. Baudin, 2004, Enhanced trapping of molybdenum by
sulfurized marine organic matter of marine origin in Mesozoic limestones and shales: Chemical
Geology, v. 213, no. 4, p. 385–401.
Tyson, R. V, 2005, The “Productivity Versus Preservation” Controversy: Cause, Flaws, And Resolution:
SEPM Special Publication No. 82, no. 82, p. 17–33.
Tyson, R. V., and T. H. Pearson, 1991, Modern and ancient continental shelf anoxia: an overview:
Geological Society, London, Special Publications, v. 58, no. 58, p. 1–24.
De Vleeschouwer, D., M. Crucifix, N. Bounceur, and P. Claeys, 2014, The impact of astronomical forcing
on the Late Devonian greenhouse climate: Global and Planetary Change, v. 120, p. 65–80.
Webster, R. L., 1984, Petroleum source rocks and stratigraphy of the Bakken Formation in North Dakota:
in Woodward J., F. F. Meissner, J. L. Clayton, eds., Hydrocarbon source rocks of the greater Rocky
Mountain region, Rocky Mountain Association of Geologists, p. 57 -82.
Wedepohl, B. K. H., 1971, Environmental Influences on the Chemical Composition of Shales and Clays:
Physics and Chemistry of the Earth, v. 8, p. 307–333.
Wedepohl, K. H., 2004, The Composition of Earth's Upper Crust, Natural Cycles of Elements, Natural
Resources, in Elements and Their Compounds in the Environment: E. Merian, M. Anke, M. Ihnat, M.
Stoeppler, eds., Occurrence, Analysis and Biological Relevance, Second Edition, Wiley-VCH Verlag
GmbH, Weinheim, Germany.
Werne, J. P., B. B. Sageman, T. W. Lyons, and D. J. Hollander, 2002, An Integrated Assessment of a “Type
Euxinic” deposit: Evidence for multiple controls on Black shale deposition in the Middle Devonian
Oatka Creek Formation: American Journal of Science, v. 302, p. 110–143.
De Wever P, P. Dumitrica, J. P., Caulet, C. Nigrini C, M. Caridroit, 2002, Radiolarians in the sedimentary
record. Amsterdam: Gordon Breach Science Publishers, p. 533.

157
Zheng, Y., R. F. Anderson, A. Van Geen, and M. Q. Fleisher, 2002, Preservation of particulate non-
lithogenic uranium in marine sediments: Geochimica et Cosmochimica Acta, v. 66, no. 17, p. 3085–
3092.
Zheng, Y., R. F. Anderson, A. Van Geen, and J. Kuwabara, 2000, Authigenic molybdenum formation in
marine sediments: A link to pore water sulfide in the Santa Barbara Basin: Geochimica et
Cosmochimica Acta, v. 64, no. 24, p. 4165–4178.

158
 5. CHAPTER 5

CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK

The purpose of this thesis is to improve the current geological understanding of the Bakken
Formation. The three primary objectives of this thesis were as follows: 1) to develop an understanding of
the pore-types, pore-architecture, pore-size, and the role of dolomitization on porosity evolution in the
Middle Bakken Member (MBM) of the Elm Coulee Field; 2) to identify the lithofacies of the Upper Bakken
Shale (UBS) and their lateral and vertical distribution, define laterally correlatable chemostratigraphic units
of UBS, their depositional conditions, and their thickness variation in the basin; and 3) to develop an
understanding of the interplay of paleoproductivity, redox condition and dilution of OM in controlling the
organic-richness of the UBS and develop a conceptual model for its deposition. These research objectives
were attained by integrating multiple data-set and methods, which include core-description, thin-section,
and SEM petrography, MIP, routine-core analysis data, XRF, XRD, ICP-MS, stable isotope of carbon,
oxygen, and nitrogen, and TOC obtained from SRA. The conclusions of the thesis are summarized in
Section 0; the recommendations for future work are listed in Section 5.2.

5.1 Conclusion

The general findings of this research are summarized in following five-points:

1) Dolomitization of the MBM in the Elm Coulee Field has taken place by the seepage-reflux of
mesohaline to penesaline brine. As a major diagenetic event, the process of dolomitization has
played a significant role in the porosity evolution of the MBM of the Elm Coulee Field. Early
patchy calcite cement, anhydrite cement, and authigenic illite clay are the other diagenetic factors
that reduce the porosity of the Middle Bakken.
2) The MBM in the Elm Coulee Field consists of sub-micron sized intercrystalline pores between the
dolomite rhombs. Total porosity and pore throat size increases as dolomite content increases up to
60 wt.%. When dolomite content increases more than 60 wt.%, total porosity and pore throat size
start decreasing due to overdolomitization. Moreover, pore-bridging and pore-lining effect of the
authigenic illite clay within the intercrystalline pores between the dolomite rhombs reduces the
pore throat size and forms a complex pore network in the MBM.
3) Six different lithofacies (F-1 through F6) were identified in the UBS; however, 95% of the UBS
primarily consists of facies F-1, F-2, and F-4. Basin-wide variation in the lateral and vertical
distribution of these facies was observed.

159
 4) Three laterally correlatable chemostratigraphic packages —sub-unit-1a, sub-unit 1b and Unit-2—
were identified in UBS. Reducing conditions developed at the sediment-water interface and in the
bottom waters due to degradation of organic matter (OM). Sub-unit 1a and 1b were deposited in a
persistently euxinic condition, which existed in the bottom water and extended to the photic zone
intermittently. However, Unit-2 was deposited in a less reducing condition, which varied from sub-
oxic to intermittently euxinic. Unit-1 was deposited during an overall transgression and is a part of
the transgressive systems tract (TST), while Unit-2 was deposited during a regression and
represents the highstand systems tract (HST). Sub-units 1a and 1b have more biogenic and
authigenic silica compared to Unit-2, which makes them comparatively more brittle than Unit-2.
5) The organic-richness of the UBS was controlled by paleoproductivity, bottom-water redox
conditions, and dilution by detrital silt-sized sediments and radiolarians. For the different
chemostratigraphic sub-units of the UBS, the effect of these factors varied, primarily due to a
relative change in the sea level. This has resulted in a variation in the organic-richness and OM
type of the chemostratigraphic sub-units of the UBS. Paleoproductivity was high during the
deposition of Unit-1 as more biolimiting nutrients (P, N) were getting preferentially regenerated
due to the euxinic paleoredox condition. This resulted in high TOC in the lower Unit-1 compared
to that of the upper Unit-2 of the UBS in the southern part of the basin depocenter.

5.2 Recommendations for Future Work

1) The dolomitization model for the Elm Coulee Field proposed in this study should be compared with
other parts of the basin. This will give an idea of the regional flow pattern of the dolomitizing fluid,
which in turn will help to predict areas with better reservoir quality.
2) Stable isotope data for sulfur should be acquired from the MBM of the Elm Coulee Field samples,
which consists of secondary anhydrite nodules and cement. This will give a better idea about the
timing of the precipitation of anhydrite cement. Stable isotope of C and O from samples, which
consists of patchy calcite cement, should be acquired to understand the genesis of calcite cement.
3) In this research, cathodoluminescence study was conducted on petrographic thin sections of the
MBM; however, the petrographic thin sections were not optimally polished. Different dolomite
zones indicating various stages of dolomitization can be further resolved by optimally polishing
the thin section before studying them under a cathodoluminescence microscope.
4) In this study, the macrofossil-bearing silt-rich mudstone (F4) was found localized in the southern
depocenter of the Williston Basin. Additional cores from the basin depocenter, north-western and
eastern areas should be studied to verify the presence or absence of facies F4 in other parts of the

160
 basin. This will help in developing a better regional understanding of the depositional setting of the
UBS.
5) To further understand the relative hydrocarbon resource potential of the chemostratigraphic
packages of the UBS, geomechanical and rock-physical properties of the different
chemostratigraphic packages of the UBS should be studied. This will also help in predicting the
extent of natural fracturing and the ease of performing hydraulic fracturing in these packages.
Moreover, the porosity of these chemostratigraphic packages should also be characterized using
SEM studies and nitrogen adsorption experiments to understand the pore architecture and pore
connectivity.
6) In this study, the stable isotope data for organic carbon indicated the presence of type-III or
terrestrial OM, in the upper sub-units (2a, 2b) of the UBS in the southern part of the basin
depocenter. This new finding should be confirmed by a detailed organic petrographic study as this
finding has implications on resource potential and regional hydrocarbon charging pattern in the
producing intervals.
7) The methodology used in this thesis for understanding the stratigraphic framework, depositional
conditions and organic-richness of UBS should be used to characterize the Lower Bakken Shale
(LBS).

161
6. APPENDIX A: DIGITIZED CORE DESCRIPTION OF MIDDLE BAKKEN
MEMBER, ELM COULEE FIELD

This appendix includes the core-description panel for all the well-cores used in the Chapter-1.

162
163
164
165
166
167
168
169
 7. APPENDIX B: RAW DATA AND SOFT COPIES

In this appendix, the description of the files for XRF, XRD, MICP, TOC data, stable isotope, routine
core analysis (RCA) data, chemostratigraphic profiles and correlations are arranged in a tabular format.
These data for each chapter of this thesis are listed in the table, along with their directory location.
Moreover, the file details of the published papers and conference presentation, and additional data collected
during the PhD curriculum are also provided.

Table B.1: Table listing the file description for data used in Chapter-2 of the thesis.
File name Description Directory location
XRD and porosity data for Middle Bakken
XRD_porosity.xlsx H:\Dipanwita Thesis\Chapter-2
interval in eight well of Elm Coulee Field
Stable isotope data for carbon and oxygen of
Stable isotope.xlsx Middle Bakken Interval from four wells of Elm H:\Dipanwita Thesis\Chapter-2
Coulee Field
Micromeretics_Elm
Coulee_Drainage
only MICP.xlsx
121064 HPMI Well
Foghorn-Irvin 20-3-
Mercury intrusion porosimetry data for pore-
HLID3.xlsx
throat size distribution of four wells of Elm H:\Dropbox\thesis\Chapter-2\MICP
121064 HPMI Well
Coulee Field
Bullwinkle-Yahoo 4-
1-HSU
121064 HPMI Well
Foghorn-Irvin 20-3-
HLID3

Table B.2: Table listing the file description for the data used in Chapter-3 of the thesis.
File name Description Directory location
Compilation_XRF.xlsx XRF data from UBS of nineteen wells H:\Dipanwita Thesis\Chapter-3
collected by the author
TOC.xlsx TOC data from UBS of two wells collected by H:\Dipanwita Thesis\Chapter-3
the author and available from external sources
XRD_Chapter 3.xlsx XRD data from UBS collected by the author
H:\ Dipanwita Thesis \Chapter-3
and available from external sources
All
chemostratigraphic
profiles.pdf
Chemostratigraphic profiles for sixteen wells
Chemostratigraphic H:\ Dipanwita Thesis \Chapter-3
and correlation panels
correlation I.jpeg
Chemostratigraphic
correlation II.jpeg

170
Table B.3: Table listing the file description for the data used in Chapter-4 of the thesis.
File name Description Directory location
ICP-MS and Stable ICP-MS elemental concentration data and SRA H:\ Dipanwita Thesis \Chapter 4
Isotope data.xlsx TOC for UBS in RT well

Table B.4: Table listing the file description of additional data of the Upper and Lower Bakken Shale
collected during the PhD curriculum.
File name Description Directory location
C and N isotope Stable isotope of C and N of UBS for well LT H:\ Dipanwita Thesis \Additional
Lonetree Edna.xlsx data
S isotope Lonetree Stable isotope of S of UBS for well LT H:\ Dipanwita Thesis \Additional
Edna.xlsx data
ICP-MS Lonetree ICP-MS of UBS for well LT H:\ Dipanwita Thesis \Additional
Edna.xlsx data
XRF_LBS XRF data of Lower Bakken Shale and H:\ Dipanwita Thesis \Additional
Pronghorn Member for eight wells data

Table B.5: Table listing the file description for selected published papers and conference presentations
during the PhD curriculum.
File name Description Directory location
URTEC.pdf Application of Inorganic Geochemical Studies H:\ Dipanwita Thesis \Additional
for Characterization of Bakken Shales, data
Williston Basin, North Dakota and Montana
MB-D Mixed Siliciclastic-Carbonate System of “D” H:\ Dipanwita Thesis \Additional
Facies in the Bakken Formation, Williston data
Basin

171
8. APPENDIX C: XRF MEASUREMENT QUALITY ASSURANCE AND QUALITY
CHECK

Chapter-3 of this thesis used elemental concentration data for chemostratigraphic analysis of the
Upper Bakken Shale (UBS). XRF measurements were performed in this study using the Niton XL3t
GOLDD+ handheld energy dispersive x-ray fluorescence (ED-XRF) analyzer in the TAG mode. In this
appendix, procedures for the quality control and quality check (QA/QC) of these XRF measurements are
described. Firstly, the samples used for QA/QC evaluation and their preparation steps are described.
Thereafter, the evaluation steps and results for the optimum time required for achieving the maximum
precision for XRF measurements are elaborated upon. Finally, the steps and results for XRF measurement
reproducibility and repeatability are described.

8.1 Sample Preparation

Thirteen core-chips were collected from the core Roberts Trust 1-13H (RT) from the North Dakota
Geological Survey (NDGS) Core Repository Center in Grand Forks, ND. The location, API, and other
details of the core RT are given in Table 3.1. The core-chips were cleaned with deionized water to remove
salt crust and/or drilling mud residues, and subsequently, the samples were air dried for 24 hours. The
cleaned samples were crushed by mortar and pestle to 200-mesh size powder. About 5g aliquots of the
powdered samples were used for determining major, minor and trace element concentrations by SGS
Mineral Services, Canada. This involved performing ICP-AES and ICP-MS analyses after sodium peroxide
fusion of the powdered aliquots (SGS method GE ICM90A). Around 15g aliquots of the powdered samples
were used for making UBS matrix-specific standard cups, which is shown in Figure C.1, using the
methodology described by Hall et al. (2011). The inventory description of the standard cups used in this
study is as follows: ThermoPlastic® Seal Venting, External Overflow Reservoirs, Single Open End, 1.22"
O.D. X 0.87" High. The standard cups were filled with powdered aliquots and covered with Mylar™ sheets,
which allowed multiple uses by preventing spilling and contamination. The inventory details of the Mylar™
sheets used in this study are as follows: Mylar® 2.5μm 3.0" SpectroMembrane® Thin-Film Sample
Supports, 0.1 mm thick, 76.2mm perforation.

172
Figure C.1: Photograph showing two of the thirteen UBS-matrix specific sample standard cups, which were
used in this study for XRF measurement QA-QC purpose.

8.2 Optimal Detection Time

The precision of element concentration measurements using handheld XRF instruments depends
on detection time. In general, longer detection times give more precise measurements. The handheld XRF
instrument used in this study reports the measurement error as two standard deviations (2σ), which is
equivalent to a 95% confidence interval. The optimum detection time for the sample-type (organic-rich
shale samples in case of this study) is obtained by repeating measurements and gradually increasing the
detection time. Both absolute measurements and error (2σ) show asymptotic trends with detection time,
and at the optimum time, the absolute concentration values get stabilized, and error decreases to residual
error or the instrument’s detection limit. Moreover, by appropriately changing the detection time (DT) for
four different excitation filters (Main, Low, High, and Light), the detection precision for various elements
can be optimized in the Niton XRF. Optimizing the Light filter DT increases the detection precision of Mg
(atomic number 12) through Cl (17), whereas the measurement precision of K (19) up to Cr (24) is
optimized by changing the Low filter DT. The Main filter DT is used to optimize the measurement precision
of Mn (25) through Bi (83), whereas the changing the High filter DT optimizes the detection precision of
Ag (47) through Ba (56) (Thermo Fisher Scientific, 2014). Kocman (2014) used the Niton ED-XRF, which
was also used in this study, for XRF measurements of the samples from the Bakken Formation and
suggested that correlative trends are achieved at both 75 seconds (Main-15s, Low-15s, High-15s, Light-
30s) and 120s (Main-30s, Low-30s, High-30s, Light-30s). Nakamura (2015) used the same instrument to
study the Niobrara Formation and suggested that the optimum detection time for the Niobrara samples is
180s (Main: 30s, Low: 30s; High: 30s and Light: 90s).

The optimal DT for four excitation filters for UBS samples was determined by performing multiple
measurements on the UBS matrix-specific standard cup RT-UBS-20. As shown in Figure C.2, these
measurements involved following four different excitation filters sequences (EFS): 1) Main-15s, Low-15s,
173
High-15s, Light-30sec (Total DT-75s); 2) Main-30s, Low-30s, High-30s, Light-30sec (Total DT-120s); 3)
Main-30s, Low-30s, High-30s, Light-90sec (Total DT-180s); and 4) Main-60s, Low-60s, High-60s, Light-
90s (Total DT-270s). Five measurements were taken for each of these four EFS; therefore, in total twenty
measurements were taken on the RT-UBS-20 standard cup. In this study, the measurement precision for
each element is represented by the relative standard deviation, (σ⁄μ) or RSD, which allows the comparison
of measurement precision for different elements used in this study. The results for the DT optimization
analysis are shown in Figure C.3, which includes plots for absolute concentration and RSD for elements Si,
Al, K, Mo, U, and V. Figure C.3 shows that for EFS-2 (total DT-120s), the absolute concentration is higher,
and RSD is smaller for all elements except for light elements Al and Si, in comparison to that in EFS-1
(total DT-75s). However, for EFS-3 (total DT-180s), the absolute concentration is higher, and RSD is
smaller for Al and Si, in comparison to that in EFS-2 (total DT-120s). Finally, for EFS-4 (total DT-270s),
both absolute concentration and RSD show little change in comparison to that in EFS-3 (total DT-180s).
The significance of these results is illustrated in Figure C.2, which suggests that the optimum DT for K,
Mo, U and V is 120s and that for light elements Si and Al is 180s. Therefore, this analysis suggests that the
optimum DT for UBS is 180s.

Figure C.2: Optimal detection time for the Main, Low, High and Light filters for UBS is shown
schematically. The effect of an increase in the detection duration on the measurement of absolute
concentration and the error ( �) also shown.

174
Figure C.3: The change in elemental concentration measurement and relative error for different excitation
filters sequences (EFS) for Si, Al, K, Mo, U, and V is shown. Optimum time and EFS is characterized by
stabilization of elemental concentration readings and relative error values. For light elements Si and Al, the
optimum DT for the UBS samples with Niton handheld XRF were 180s and that for other elements was
120s.

8.3 Reproducibility Test and XRF Measurement Validation

Niton measurements were validated by comparing XRF measurements with inductively coupled
plasma mass spectrometry (ICP-MS) measurements. ICP-MS measurements were performed on all thirteen
aliquots of powder, with which the thirteen UBS standard cups were made. XRF measurements were
performed on all thirteen UBS standard cups at EFS-3 (Total DT 180s) in TAG mode. Theoretically, the
concentration of elements measured by two methods are termed reproducible, if the straight line-fit in the
crossplot between the measurements have the regression-fit coefficient (R2) and slope (m) close to one, and
the y-intercept close to zero. Therefore, in this study the elements for which, R2 > 0.9, 0.9 < m < 1.1 and
c=0, XRF data were considered as suitable for both quantitative and qualitative analyses. Elements for
which, 0.75< R2 < 0.9, and 0.8 < m < 1.2, XRF data were used for semi-quantitative and qualitative analyses.
Finally, elements for which, 0.6 < R2 < 0.75, 0.5 < m < 2.0, XRF data were used only for qualitative
analyses. Figure C.4 shows the crossplot between these XRF and ICP-MS measurements on thirteen UBS
standard cups for elemental concentrations of major and trace elements. Table C.1 summarizes the results
of the straight-line regression fitting on the crossplots for elements and their analyses applications
(quantitative, semi-quantitative, and qualitative) in this study. For qualitative analyses of the XRF data, the
elements with a positive correlation and R2 above 0.6 were observed to be suitable for trend reproducibility,
175
as it will be later shown. Moreover, as shown in Figure C.5, the log view shows similar stratigraphic trends
between the ICP-MS and XRF measurements, which further validates the applicability of XRF
measurements.

Figure C.4: Niton-ICP/Element Analyzer cross plots for eighteen selected elements showing the
relationships between Niton measured elemental concentrations and laboratory-grade measurements made
by ICP-MS/ICP-OES and element analyzer. Best-fit linear regression lines are shown with associated R2
values for each plot. For elements Al, Si, Ca, Fe, S, Mo, and U have a zero y-axis intercept.

176
Table C.1: Table for the coefficient of regression (R2) and slope (m) for the best-fit linear regression lines
for the Niton-ICP/Element Analyzer cross plots for eighteen selected elements showing the relationships
between Niton measured elemental concentrations and laboratory-grade measurements made by ICP-
MS/ICP-OES and element analyzer. The XRF-measured concentration of elements which has been used
quantitatively and semi-quantitatively in Chapter 3 are highlighted in red.
Element R2 Slope/m
Mg 0.66 0.3
Al 0.67 0.65
Si 0.71 0.87
S 0.87 0.9
K 0.83 0.62
Ca 0.98 1
Ti 0.6 0.63
Fe 0.96 1.32
U 0.95 1.28
V 0.85 0.79
Mo 0.99 1.18
Ni 0.92 0.52
Cu 0.91 0.65
Zn 0.99 0.64
Sr 0.99 0.91
Mn 0.71 0.67
Rb 0.75 0.42
Th 0.57 1.21

Figure C.5: Profile of the representative suite of major and trace elements comparing Niton XRF and ICP-
MS measurements. Visual identification of the elemental trend between the Niton XRF and ICP-MS
measurement show consistency.
177
8.4 Repeatability Test

The repeatability of the XRF measurements was analyzed by following the procedure outlined by
Fisher et al. (2014) and Nakamura (2015). A repeatability test was performed at eleven UBS core samples
from well RT, and at each depth, measurements were repeated five times with the detection time of 180
seconds. The procedures for taking XRF measurement in TAG mode are described in Section 3.3.2. Five
measurements at sample depth x were performed in different test-runs, such that each run includes
measurements at all eleven depths before repeating measurements at sample depth x. Table C.2 lists the
concentration of trace elements (Mo, U, Ni and V) and major elements (Si, Al, K, Ca, Fe, Ca, S, Ti), which
were used in this study for all five runs on each of the eleven samples (represented by their depth). The
result of the repeatability test is shown in Figure C.6, which plots depth profiles of concentration for trace
and major elements for eleven UBS samples from well RT for each test runs. Repeatability of the XRF
measurement using Niton handheld XRF in TAG mode is confirmed by Figure C.6, which shows
consistency in both major inflections and concentration changes for five test-runs.

Table C.2: The results for element concentration measurements using Niton handheld XRF for the
repeatability test is listed. The table lists the results for the concentration of trace elements and major
element, which were used in this study, for all five runs for each of the eleven samples (represented by their
depth).
Test
Depth Mo U Ni Fe V Ti Ca K S Al Si Mg
Run
ft ppm ppm ppm % ppm % % % % % % %
Run-1 8.34 17.5 127.3 2.00 634.35 0.29 7.11 2.48 0.72 3.71 20.91 0.66
Run-2 7.27 14.9 142.1 2.02 626.92 0.29 7.08 2.48 0.73 3.80 20.97 0.97
10651 Run-3 8 16.6 147.8 2.02 673.24 0.29 7.11 2.45 0.72 3.83 21.06 0.69
Run-4 6.21 18.4 149.1 2.00 646.12 0.28 7.14 2.48 0.73 3.86 21.14 0.87
Run-5 6.34 18.2 153.1 2.05 663.46 0.29 7.12 2.44 0.72 3.87 21.10 0.78
Run-1 44.53 20.9 214.1 2.30 1209.59 0.32 5.40 2.71 1.02 4.26 22.21 0.85
Run-2 45.66 22.1 222.5 2.35 1223.69 0.32 5.40 2.70 1.01 4.31 22.32 0.75
10651.5 Run-3 48.12 22.1 234.6 2.35 1187.89 0.32 5.39 2.69 1.02 4.24 22.25 0.74
Run-4 46.63 22.6 212.7 2.34 1221.04 0.32 5.37 2.69 1.01 4.34 22.23 0.92
Run-5 46.74 22.0 221.0 2.34 1227.70 0.32 5.38 2.74 1.01 4.32 22.43 0.78
Run-1 20.11 18.2 208.7 1.91 877.79 0.32 7.15 2.48 0.67 4.45 23.36 1.02
Run-2 19.88 16.8 200.6 1.88 896.09 0.32 7.12 2.50 0.68 4.42 23.41 1.02
10655.75 Run-3 19.04 18.6 230.3 1.89 930.41 0.33 7.13 2.48 0.68 4.42 23.43 1.13
Run-4 19.84 21.3 194.1 1.89 927.55 0.33 7.09 2.46 0.67 4.45 23.52 0.99
Run-5 19.18 17.8 190.5 1.90 866.29 0.33 7.13 2.49 0.67 4.47 23.45 1.06
Run-1 23.21 24.5 174.6 2.42 716.24 0.26 5.14 2.34 1.04 3.48 22.32 0.61
10658
Run-2 22.08 25.9 175.5 2.41 721.01 0.26 5.15 2.32 1.03 3.51 22.33 0.61

178
Table C.2 continued

Run-3 21.27 24.8 166.0 2.41 716.45 0.26 5.17 2.34 1.04 3.51 22.29 0.56
Run-4 21.39 25.8 151.4 2.41 718.94 0.26 5.20 2.32 1.04 3.54 22.41 0.79
Run-5 21.91 23.6 164.5 2.40 687.92 0.26 5.18 2.31 1.04 3.51 22.41 0.74
Run-1 55.2 39.1 166.2 5.05 292.63 0.22 2.11 1.82 4.34 3.59 32.29 0.78
Run-2 54.58 37.3 158.0 5.00 292.40 0.21 2.09 1.80 4.34 3.59 32.29 0.78
10661 Run-3 53.38 40.6 167.0 5.01 319.58 0.22 2.12 1.82 4.35 3.58 32.44 0.96
Run-4 54.15 35.6 173.1 4.99 304.12 0.21 2.10 1.80 4.34 3.44 32.33 0.53
Run-5 54.16 32.3 150.7 4.95 317.09 0.21 2.10 1.81 4.35 3.54 32.35 0.60
Run-1 323.05 100.8 272.4 3.34 868.90 0.29 1.05 2.20 2.71 3.96 30.50 0.72
Run-2 322.41 105.7 276.9 3.34 888.83 0.29 1.07 2.23 2.71 3.93 30.61 0.64
10663 Run-3 322.39 100.8 292.8 3.35 875.26 0.30 1.08 2.25 2.73 4.05 30.68 0.60
Run-4 318.94 98.3 281.6 3.33 900.32 0.29 1.07 2.26 2.72 4.05 30.65 0.76
Run-5 324.2 109.7 283.6 3.33 875.58 0.30 1.09 2.25 2.72 4.04 30.68 0.62
Run-1 243.88 44.6 170.8 2.39 362.20 0.21 0.74 1.84 1.46 2.40 26.58 0.39
Run-2 238.65 47.5 149.2 2.41 378.94 0.21 0.74 1.85 1.46 2.49 26.79 0.48
10664 Run-3 241.08 47.7 143.9 2.41 382.06 0.20 0.73 1.85 1.48 2.50 26.90 0.43
Run-4 239.86 49.1 171.6 2.39 370.51 0.22 0.73 1.84 1.46 2.50 26.93 0.27
Run-5 240.98 43.7 159.2 2.41 381.38 0.21 0.73 1.85 1.47 2.56 27.15 0.37
Run-1 348.19 60.5 207.4 2.76 775.64 0.24 0.81 1.82 2.39 2.91 31.20 0.56
Run-2 345.03 67.4 206.9 2.81 770.98 0.24 0.80 1.84 2.41 2.95 31.31 0.57
10666.5 Run-3 346.85 66.5 209.0 2.78 779.99 0.24 0.81 1.81 2.42 2.93 31.40 0.50
Run-4 350.07 68.8 198.3 2.77 754.33 0.24 0.81 1.79 2.42 2.98 31.56 0.48
Run-5 340.77 68.4 203.7 2.76 774.08 0.24 0.81 1.81 2.41 2.94 31.49 0.50
Run-1 412.73 79.8 337.9 3.37 1086.96 0.26 0.84 2.20 2.63 3.42 28.54 0.56
Run-2 404.68 82.3 329.2 3.38 1111.89 0.25 0.84 2.20 2.64 3.44 28.61 0.54
10667.5 Run-3 415.42 76.4 315.5 3.34 1090.23 0.26 0.85 2.21 2.63 3.44 28.52 0.57
Run-4 409.56 79.6 322.9 3.33 1101.34 0.26 0.84 2.20 2.63 3.46 28.62 0.67
Run-5 409.56 74.3 321.3 3.35 1088.88 0.25 0.83 2.19 2.64 3.46 28.73 0.51
Run-1 328.11 62.6 317.2 3.76 1154.68 0.30 1.59 2.29 3.00 3.76 26.52 0.47
Run-2 331.93 70.2 318.1 3.77 1156.46 0.30 1.57 2.28 3.00 3.71 26.56 0.67
10668.5 Run-3 333.07 70.7 331.9 3.77 1149.81 0.30 1.60 2.27 3.02 3.84 26.75 0.84
Run-4 333.58 67.1 314.7 3.79 1138.83 0.29 1.59 2.32 3.05 3.91 26.78 0.75
Run-5 340.26 66.0 328.7 3.78 1149.19 0.29 1.60 2.28 3.05 3.81 26.75 0.56
Run-1 373.44 83.0 243.5 3.98 649.78 0.27 1.08 2.35 3.19 3.63 25.07 0.67
Run-2 368.2 84.2 231.6 3.96 642.15 0.26 1.09 2.37 3.17 3.70 24.96 0.64
10669.5 Run-3 369.28 84.2 203.9 3.97 645.70 0.27 1.07 2.35 3.19 3.62 25.10 0.57
Run-4 367.41 89.3 212.9 3.98 655.54 0.27 1.08 2.39 3.19 3.63 25.22 0.48
Run-5 378.59 87.4 221.0 3.97 666.34 0.26 1.10 2.41 3.21 3.70 25.34 0.54

179
Figure C.6: Depth profiles of concentration show the repeatability test results for trace and major elements for eleven UBS samples (represented by
sample depth) from well RT for each test runs. The error bars indicate the variability in elemental concentration measurement is for five test run for
a sample of given depth. The red line connects the average of the five runs, whereas the blue line shows the spread of the measurement. Repeatability
of the XRF measurement using Niton handheld XRF in TAG mode is confirmed by consistency in both major inflections and concentration changes
for five test-run.

180
 The percentage variability (�⁄�) in the measurement of concentration an element for a sample in
five repeat-runs was determined by normalizing the standard deviation (�) of the reading with average �.
The variability results are shown in Figure C.7 in a heat-map format, in which the cooler color (green) and
lower percentages indicates better repeatability. The last row in the heat-map in Figure C.7 shows the
average of the variability for the eleven UBS samples from well RT, which shows that the measurement of
trace element using Niton handheld XRF is repeatable within an average range of 2-5%, whereas for major
elements, repeatability is within 0.2-2%.

Depth (ft) Mo U Ni V Fe Ti Ca K S Al Si Mg
10651 13.2% 8.4% 7.0% 3.0% 1.0% 1.1% 0.3% 0.7% 0.6% 1.6% 0.5% 16.0%
10651.5 2.9% 2.8% 3.9% 1.3% 0.8% 0.9% 0.2% 0.7% 0.4% 1.0% 0.4% 9.3%
10655.75 2.4% 9.0% 7.7% 3.2% 0.6% 1.3% 0.3% 0.5% 0.6% 0.4% 0.3% 5.1%
10658 3.5% 3.9% 5.8% 1.9% 0.3% 1.1% 0.4% 0.5% 0.3% 0.6% 0.3% 14.6%
10661 1.2% 8.7% 5.4% 4.2% 0.7% 1.7% 0.5% 0.5% 21.1% 2.0% 0.2% 27.0%
10663 0.6% 4.4% 2.7% 1.4% 0.2% 1.6% 1.3% 1.0% 0.3% 1.4% 0.2% 10.5%
10664 0.8% 5.0% 7.8% 2.3% 0.4% 2.0% 0.7% 0.3% 0.5% 2.3% 0.8% 19.7%
10666.5 1.0% 5.1% 2.1% 1.3% 0.7% 0.7% 0.6% 1.1% 0.6% 0.8% 0.5% 7.5%
10667.5 1.0% 4.0% 2.6% 1.0% 0.6% 2.1% 0.9% 0.3% 0.2% 0.6% 0.3% 10.4%
10668.5 1.3% 4.9% 2.4% 0.6% 0.3% 1.4% 0.6% 0.8% 0.8% 2.0% 0.5% 22.0%
10669.5 1.3% 3.0% 7.0% 1.5% 0.2% 1.1% 1.0% 1.1% 0.4% 1.1% 0.6% 13.0%
Average 2.7% 5.4% 5.0% 2.0% 0.5% 1.3% 0.6% 0.7% 2.3% 1.3% 0.4% 14.1%

Figure C.7: Repeatability of the XRF element concentration measurements for the eleven UBS samples
from core RT is shown in the heat-map, in which the numbers represent the variation (σ⁄μ) The color-scale
and the bar show the relative degree of repeatability for element concentration measurement by XRF.
Cooler color (green) and lower percentages indicate better repeatability. Mg (Atomic No.-12), which was
not used in this study, in general, shows poor repeatability because hand-held XRF and TAG mode used in
this study is not suitable for detecting elements lighter than Al (Atomic No.-13).

181
9. APPENDIX D: CHEMOSTRATIGRAPHIC PROFILES AND CORRELATIONS

Chemostratigraphic profiles for sixteen wells and chemostratigraphic correlation for cross-sections
I and cross-section II used in Chapter-3 are included in this appendix.

182
183
184
185
186